Claus Stadler

2023

• Base Plat­form for Knowl­edge Graphs with Free Soft­wareIn: Sebas­t­ian Tramp, Ricar­do Usbeck, Natanael Arndt, Julia Holze and Sören Auer (eds.): Pro­ceed­ings of the Inter­na­tion­al Work­shop on Linked Data-dri­ven Resilience Research 2023, {CEUR} Work­shop Pro­ceed­ings. vol. 3401. Her­sonis­sos, GreeceSimon Bin, Claus Stadler, Nor­man Radtke, Kurt Jung­hanns, Sabine Grün­der-Fahrer and Michael Martin

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  We present an Open Source base plat­form for the Coy­Pu knowl­edge graph project in the resilience domain. We report on our expe­ri­ences with sev­er­al tools which are used to cre­ate, main­tain, serve, view and explore a mod­u­lar large-scale knowl­edge graph, as well as the adap­tions that were nec­es­sary to enable fric­tion­less inter­ac­tion from both per­for­mance and usabil­i­ty per­spec­tives. For this pur­pose, sev­er­al adjust­ments had to be made. We pro­vide a broad view of dif­fer­ent pro­grams which are of rel­e­vance to this domain. We demon­strate that while it is already pos­si­ble to achieve good results with free soft­ware, there are still sev­er­al pain points that need to be addressed. Res­o­lu­tion of these issues is often not only a mat­ter of con­fig­u­ra­tion but requires mod­i­fi­ca­tion of the source code as well.

  @inproceedings{bin-2023–base-platform,
  abstract = {We present an Open Source base plat­form for the Coy­Pu knowl­edge graph project in the resilience domain. We report on our expe­ri­ences with sev­er­al tools which are used to cre­ate, main­tain, serve, view and explore a mod­u­lar large-scale knowl­edge graph, as well as the adap­tions that were nec­es­sary to enable fric­tion­less inter­ac­tion from both per­for­mance and usabil­i­ty per­spec­tives. For this pur­pose, sev­er­al adjust­ments had to be made. We pro­vide a broad view of dif­fer­ent pro­grams which are of rel­e­vance to this domain. We demon­strate that while it is already pos­si­ble to achieve good results with free soft­ware, there are still sev­er­al pain points that need to be addressed. Res­o­lu­tion of these issues is often not only a mat­ter of con­fig­u­ra­tion but requires mod­i­fi­ca­tion of the source code as well.},
  address = {Her­sonis­sos, Greece},
  author = {Bin, Simon and Stadler, Claus and Radtke, Nor­man and Jung­hanns, Kurt and Grün­der-Fahrer, Sabine and Mar­tin, Michael},
  book­ti­tle = {Pro­ceed­ings of the Inter­na­tion­al Work­shop on Linked Data-dri­ven Resilience Research 2023},
  edi­tor = {Tramp, Sebas­t­ian and Usbeck, Ricar­do and Arndt, Natanael and Holze, Julia and Auer, Sören},
  key­words = {sys:relevantFor:infai},
  month = {05},
  series = {{CEUR} Work­shop Pro­ceed­ings},
  title = {Base Plat­form for Knowl­edge Graphs with Free Soft­ware},
  vol­ume = 3401,
  year = 2023
  }
  

  %0 Con­fer­ence Paper
  %1 bin-2023–base-platform
  %A Bin, Simon
  %A Stadler, Claus
  %A Radtke, Nor­man
  %A Jung­hanns, Kurt
  %A Grün­der-Fahrer, Sabine
  %A Mar­tin, Michael
  %B Pro­ceed­ings of the Inter­na­tion­al Work­shop on Linked Data-dri­ven Resilience Research 2023
  %C Her­sonis­sos, Greece
  %D 2023
  %E Tramp, Sebas­t­ian
  %E Usbeck, Ricar­do
  %E Arndt, Natanael
  %E Holze, Julia
  %E Auer, Sören
  %T Base Plat­form for Knowl­edge Graphs with Free Soft­ware
  %U https://ceur-ws.org/Vol-3401/paper6.pdf
  %V 3401
  %X We present an Open Source base plat­form for the Coy­Pu knowl­edge graph project in the resilience domain. We report on our expe­ri­ences with sev­er­al tools which are used to cre­ate, main­tain, serve, view and explore a mod­u­lar large-scale knowl­edge graph, as well as the adap­tions that were nec­es­sary to enable fric­tion­less inter­ac­tion from both per­for­mance and usabil­i­ty per­spec­tives. For this pur­pose, sev­er­al adjust­ments had to be made. We pro­vide a broad view of dif­fer­ent pro­grams which are of rel­e­vance to this domain. We demon­strate that while it is already pos­si­ble to achieve good results with free soft­ware, there are still sev­er­al pain points that need to be addressed. Res­o­lu­tion of these issues is often not only a mat­ter of con­fig­u­ra­tion but requires mod­i­fi­ca­tion of the source code as well.
  

• {KGCW}2023 Chal­lenge Report {RDF}{P}rocessing{T}oolkit / SansaIn: 4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023, CEUR work­shop pro­ceed­ings. Her­sonis­sos, GreeceSimon Bin, Claus Stadler and Lorenz Bühmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  This is the report of our par­tic­i­pa­tion in the KGCW2023 Chal­lenge @ ESWC 2023 with our RDFProcessingToolkit/Sansa sys­tem which won the “fastest” tool award. The chal­lenge was about the con­struc­tion of RDF knowl­edge graphs from RML spec­i­fi­ca­tions with vary­ing com­plex­i­ty in regard to the mix of input for­mats, char­ac­ter­is­tics of the data and the need­ed join oper­a­tions. We detail how we inte­grat­ed our tool into the pro­vid­ed bench­mark frame­work. There­by we also report on the issues and short­com­ings we encoun­tered as a base for future improve­ments. Fur­ther­more, we pro­vide an analy­sis of the data mea­sured with the bench­mark framework.

  @inproceedings{stadler2023-kgcw-challenge,
  abstract = {This is the report of our par­tic­i­pa­tion in the KGCW2023 Chal­lenge @ ESWC 2023 with our RDFProcessingToolkit/Sansa sys­tem which won the “fastest” tool award. The chal­lenge was about the con­struc­tion of RDF knowl­edge graphs from RML spec­i­fi­ca­tions with vary­ing com­plex­i­ty in regard to the mix of input for­mats, char­ac­ter­is­tics of the data and the need­ed join oper­a­tions. We detail how we inte­grat­ed our tool into the pro­vid­ed bench­mark frame­work. There­by we also report on the issues and short­com­ings we encoun­tered as a base for future improve­ments. Fur­ther­more, we pro­vide an analy­sis of the data mea­sured with the bench­mark frame­work.},
  address = {Her­sonis­sos, Greece},
  author = {Bin, Simon and Stadler, Claus and Büh­mann, Lorenz},
  book­ti­tle = {4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023},
  key­words = {sys:relevantFor:infai},
  num­ber = 3471,
  series = {CEUR work­shop pro­ceed­ings},
  title = {{KGCW}2023 Chal­lenge Report {RDF}{P}rocessing{T}oolkit / Sansa},
  year = 2023
  }
  

  %0 Con­fer­ence Paper
  %1 stadler2023-kgcw-chal­lenge
  %A Bin, Simon
  %A Stadler, Claus
  %A Büh­mann, Lorenz
  %B 4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023
  %C Her­sonis­sos, Greece
  %D 2023
  %N 3471
  %T {KGCW}2023 Chal­lenge Report {RDF}{P}rocessing{T}oolkit / Sansa
  %U https://ceur-ws.org/Vol-3471/paper12.pdf
  %X This is the report of our par­tic­i­pa­tion in the KGCW2023 Chal­lenge @ ESWC 2023 with our RDFProcessingToolkit/Sansa sys­tem which won the “fastest” tool award. The chal­lenge was about the con­struc­tion of RDF knowl­edge graphs from RML spec­i­fi­ca­tions with vary­ing com­plex­i­ty in regard to the mix of input for­mats, char­ac­ter­is­tics of the data and the need­ed join oper­a­tions. We detail how we inte­grat­ed our tool into the pro­vid­ed bench­mark frame­work. There­by we also report on the issues and short­com­ings we encoun­tered as a base for future improve­ments. Fur­ther­more, we pro­vide an analy­sis of the data mea­sured with the bench­mark framework.
  

• LLM-assist­ed Knowl­edge Graph Engi­neer­ing: Exper­i­ments with Chat­G­PTLars-Peter Mey­er, Claus Stadler, Johannes Frey, Nor­man Radtke, Kurt Jung­hanns, Roy Meiss­ner, Gor­dian Dzi­wis, Kir­ill Bulert, Michael Martin

  AbstractBib­TeXEnd­NoteBib­Son­o­my

  Knowl­edge Graphs (KG) pro­vide us with a struc­tured, flex­i­ble, trans­par­ent, cross-sys­tem, and col­lab­o­ra­tive way of orga­niz­ing our knowl­edge and data across var­i­ous domains in soci­ety and indus­tri­al as well as sci­en­tif­ic dis­ci­plines. KGs sur­pass any oth­er form of rep­re­sen­ta­tion in terms of effec­tive­ness. How­ev­er, Knowl­edge Graph Engi­neer­ing (KGE) requires in-depth expe­ri­ences of graph struc­tures, web tech­nolo­gies, exist­ing mod­els and vocab­u­lar­ies, rule sets, log­ic, as well as best prac­tices. It also demands a sig­nif­i­cant amount of work. Con­sid­er­ing the advance­ments in large lan­guage mod­els (LLMs) and their inter­faces and appli­ca­tions in recent years, we have con­duct­ed com­pre­hen­sive exper­i­ments with Chat­G­PT to explore its poten­tial in sup­port­ing KGE. In this paper, we present a selec­tion of these exper­i­ments and their results to demon­strate how Chat­G­PT can assist us in the devel­op­ment and man­age­ment of KGs.

  @article{Meyer2023LLMassistedKnowledge,
  abstract = {Knowl­edge Graphs (KG) pro­vide us with a struc­tured, flex­i­ble, trans­par­ent, cross-sys­tem, and col­lab­o­ra­tive way of orga­niz­ing our knowl­edge and data across var­i­ous domains in soci­ety and indus­tri­al as well as sci­en­tif­ic dis­ci­plines. KGs sur­pass any oth­er form of rep­re­sen­ta­tion in terms of effec­tive­ness. How­ev­er, Knowl­edge Graph Engi­neer­ing (KGE) requires in-depth expe­ri­ences of graph struc­tures, web tech­nolo­gies, exist­ing mod­els and vocab­u­lar­ies, rule sets, log­ic, as well as best prac­tices. It also demands a sig­nif­i­cant amount of work. Con­sid­er­ing the advance­ments in large lan­guage mod­els (LLMs) and their inter­faces and appli­ca­tions in recent years, we have con­duct­ed com­pre­hen­sive exper­i­ments with Chat­G­PT to explore its poten­tial in sup­port­ing KGE. In this paper, we present a selec­tion of these exper­i­ments and their results to demon­strate how Chat­G­PT can assist us in the devel­op­ment and man­age­ment of KGs.},
  author = {Mey­er, Lars-Peter and Stadler, Claus and Frey, Johannes and Radtke, Nor­man and Jung­hanns, Kurt and Meiss­ner, Roy and Dzi­wis, Gor­dian and Bulert, Kir­ill and Mar­tin, Michael},
  key­words = {sys:relevantFor:infai},
  title = {LLM-assist­ed Knowl­edge Graph Engi­neer­ing: Exper­i­ments with Chat­G­PT},
  year = 2023
  }
  

  %0 Jour­nal Arti­cle
  %1 Meyer2023LLMassistedKnowledge
  %A Mey­er, Lars-Peter
  %A Stadler, Claus
  %A Frey, Johannes
  %A Radtke, Nor­man
  %A Jung­hanns, Kurt
  %A Meiss­ner, Roy
  %A Dzi­wis, Gor­dian
  %A Bulert, Kir­ill
  %A Mar­tin, Michael
  %D 2023
  %R 10.48550/ARXIV.2307.06917
  %T LLM-assist­ed Knowl­edge Graph Engi­neer­ing: Exper­i­ments with Chat­G­PT
  %X Knowl­edge Graphs (KG) pro­vide us with a struc­tured, flex­i­ble, trans­par­ent, cross-sys­tem, and col­lab­o­ra­tive way of orga­niz­ing our knowl­edge and data across var­i­ous domains in soci­ety and indus­tri­al as well as sci­en­tif­ic dis­ci­plines. KGs sur­pass any oth­er form of rep­re­sen­ta­tion in terms of effec­tive­ness. How­ev­er, Knowl­edge Graph Engi­neer­ing (KGE) requires in-depth expe­ri­ences of graph struc­tures, web tech­nolo­gies, exist­ing mod­els and vocab­u­lar­ies, rule sets, log­ic, as well as best prac­tices. It also demands a sig­nif­i­cant amount of work. Con­sid­er­ing the advance­ments in large lan­guage mod­els (LLMs) and their inter­faces and appli­ca­tions in recent years, we have con­duct­ed com­pre­hen­sive exper­i­ments with Chat­G­PT to explore its poten­tial in sup­port­ing KGE. In this paper, we present a selec­tion of these exper­i­ments and their results to demon­strate how Chat­G­PT can assist us in the devel­op­ment and man­age­ment of KGs.
  

• Scal­ing RML and SPAR­QL-based Knowl­edge Graph Con­struc­tion with Apache SparkIn: 4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023, CEUR work­shop pro­ceed­ings. vol. 3471. Her­sonis­sos, GreeceClaus Stadler, Lorenz Büh­mann, Lars-Peter Mey­er and Michael Martin

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Approach­es for the con­struc­tion of knowl­edge graphs from het­ero­ge­neous data sources range from ad-hoc scripts to ded­i­cat­ed map­ping lan­guages. Two com­mon foun­da­tions are there­by RML and SPARQL. So far, both approach­es are treat­ed as dif­fer­ent: On the one hand there are tools specif­i­cal­ly for pro­cess­ing RML where­as on the oth­er hand there are tools that extend SPARQL in order to incor­po­rate addi­tion­al data sources. In this work, we first show how this gap can be bridged by trans­lat­ing RML to a sequence of SPARQL CONSTRUCT queries and intro­duce the nec­es­sary SPARQL exten­sions. In a sub­se­quent step, we employ tech­niques to opti­mize SPARQL query work­loads as well as indi­vid­ual query exe­cu­tion times in order to obtain an opti­mized sequence of queries with respect to the order and unique­ness of the gen­er­at­ed triples. Final­ly, we present a cor­re­spond­ing SPARQL query exe­cu­tion engine based on the Apache Spark Big Data frame­work. In our eval­u­a­tion on bench­marks we show that our approach is capa­ble of achiev­ing RML map­ping exe­cu­tion per­for­mance that sur­pass­es the cur­rent state of the art.

  @inproceedings{stadler2023-scaling-rml,
  abstract = {Approach­es for the con­struc­tion of knowl­edge graphs from het­ero­ge­neous data sources range from ad-hoc scripts to ded­i­cat­ed map­ping lan­guages. Two com­mon foun­da­tions are there­by RML and SPARQL. So far, both approach­es are treat­ed as dif­fer­ent: On the one hand there are tools specif­i­cal­ly for pro­cess­ing RML where­as on the oth­er hand there are tools that extend SPARQL in order to incor­po­rate addi­tion­al data sources. In this work, we first show how this gap can be bridged by trans­lat­ing RML to a sequence of SPARQL CONSTRUCT queries and intro­duce the nec­es­sary SPARQL exten­sions. In a sub­se­quent step, we employ tech­niques to opti­mize SPARQL query work­loads as well as indi­vid­ual query exe­cu­tion times in order to obtain an opti­mized sequence of queries with respect to the order and unique­ness of the gen­er­at­ed triples. Final­ly, we present a cor­re­spond­ing SPARQL query exe­cu­tion engine based on the Apache Spark Big Data frame­work. In our eval­u­a­tion on bench­marks we show that our approach is capa­ble of achiev­ing RML map­ping exe­cu­tion per­for­mance that sur­pass­es the cur­rent state of the art.},
  address = {Her­sonis­sos, Greece},
  author = {Stadler, Claus and Büh­mann, Lorenz and Mey­er, Lars-Peter and Mar­tin, Michael},
  book­ti­tle = {4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023},
  key­words = {sys:relevantFor:infai},
  series = {CEUR work­shop pro­ceed­ings},
  title = {Scal­ing RML and SPAR­QL-based Knowl­edge Graph Con­struc­tion with Apache Spark},
  vol­ume = 3471,
  year = 2023
  }
  

  %0 Con­fer­ence Paper
  %1 stadler2023-scal­ing-rml
  %A Stadler, Claus
  %A Büh­mann, Lorenz
  %A Mey­er, Lars-Peter
  %A Mar­tin, Michael
  %B 4th Inter­na­tion­al Work­shop on Knowl­edge Graph Con­struc­tion @ ESWC 2023
  %C Her­sonis­sos, Greece
  %D 2023
  %T Scal­ing RML and SPAR­QL-based Knowl­edge Graph Con­struc­tion with Apache Spark
  %U https://ceur-ws.org/Vol-3471/paper8.pdf
  %V 3471
  %X Approach­es for the con­struc­tion of knowl­edge graphs from het­ero­ge­neous data sources range from ad-hoc scripts to ded­i­cat­ed map­ping lan­guages. Two com­mon foun­da­tions are there­by RML and SPARQL. So far, both approach­es are treat­ed as dif­fer­ent: On the one hand there are tools specif­i­cal­ly for pro­cess­ing RML where­as on the oth­er hand there are tools that extend SPARQL in order to incor­po­rate addi­tion­al data sources. In this work, we first show how this gap can be bridged by trans­lat­ing RML to a sequence of SPARQL CONSTRUCT queries and intro­duce the nec­es­sary SPARQL exten­sions. In a sub­se­quent step, we employ tech­niques to opti­mize SPARQL query work­loads as well as indi­vid­ual query exe­cu­tion times in order to obtain an opti­mized sequence of queries with respect to the order and unique­ness of the gen­er­at­ed triples. Final­ly, we present a cor­re­spond­ing SPARQL query exe­cu­tion engine based on the Apache Spark Big Data frame­work. In our eval­u­a­tion on bench­marks we show that our approach is capa­ble of achiev­ing RML map­ping exe­cu­tion per­for­mance that sur­pass­es the cur­rent state of the art.
  

• QROWD—A Plat­form for Inte­grat­ing Cit­i­zens in Smart City Data Ana­lyt­icsIn: Pradeep Kumar Singh, Marcin Paprzy­c­ki, Mohamad Essaai­di and Shahram Rahi­mi (eds.): Sus­tain­able Smart Cities: The­o­ret­i­cal Foun­da­tions and Prac­ti­cal Con­sid­er­a­tions. Cham : Springer Inter­na­tion­al Pub­lish­ing — ISBN 978–3‑031–08815‑5, pp. 285–321Luis-Daniel Ib{á}{\~{n}}ez, Eddy Mad­dale­na, Richard Gomer, Ele­na Sim­perl, Mat­tia Zeni, Enri­co Big­not­ti, Ronald Chenu-Abente, Faus­to Giunchiglia, Patrick West­phal, et al.

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Opti­miz­ing mobil­i­ty ser­vices is one of the great­est chal­lenges Smart Cities face in their efforts to improve res­i­dents’ well­be­ing and reduce {\$}{\$}{\backslash}text {\{}CO{\}}{\_}{\{}2{\}}{\$}{\$}emissions. The advent of IoT has cre­at­ed unpar­al­leled oppor­tu­ni­ties to col­lect large amounts of data about how peo­ple use trans­porta­tion. This data could be used to ascer­tain the qual­i­ty and reach of the ser­vices offered and to inform future policy—provided cities have the capa­bil­i­ties to process, curate, inte­grate and analyse the data effec­tive­ly. At the same time, to be tru­ly ‘Smart’, cities need to ensure that the data-dri­ven deci­sions they make reflect the needs of their cit­i­zens, cre­ate feed­back loops, and widen par­tic­i­pa­tion. In this chap­ter, we intro­duce QROWD, a data inte­gra­tion and ana­lyt­ics plat­form that seam­less­ly inte­grates mul­ti­ple data sources along­side human, social and com­pu­ta­tion­al intel­li­gence to build hybrid, auto­mat­ed data-cen­tric work­flows. By doing so, QROWD appli­ca­tions can take advan­tage of the best of both worlds: the accu­ra­cy and scale of machine com­pu­ta­tion, and the skills, knowl­edge and exper­tise of peo­ple. We present the archi­tec­ture and main com­po­nents of the plat­form, as well as its usage to realise two mobil­i­ty use cas­es: esti­mat­ing the modal split, which refers to trips peo­ple take that involve more than one type of trans­port, and urban auditing.

  @inbook{Ibanez2023-qrowd,
  abstract = {Opti­miz­ing mobil­i­ty ser­vices is one of the great­est chal­lenges Smart Cities face in their efforts to improve res­i­dents’ well­be­ing and reduce {\$}{\$}{\backslash}text {\{}CO{\}}{\_}{\{}2{\}}{\$}{\$}emissions. The advent of IoT has cre­at­ed unpar­al­leled oppor­tu­ni­ties to col­lect large amounts of data about how peo­ple use trans­porta­tion. This data could be used to ascer­tain the qual­i­ty and reach of the ser­vices offered and to inform future policy—provided cities have the capa­bil­i­ties to process, curate, inte­grate and analyse the data effec­tive­ly. At the same time, to be tru­ly ‘Smart’, cities need to ensure that the data-dri­ven deci­sions they make reflect the needs of their cit­i­zens, cre­ate feed­back loops, and widen par­tic­i­pa­tion. In this chap­ter, we intro­duce QROWD, a data inte­gra­tion and ana­lyt­ics plat­form that seam­less­ly inte­grates mul­ti­ple data sources along­side human, social and com­pu­ta­tion­al intel­li­gence to build hybrid, auto­mat­ed data-cen­tric work­flows. By doing so, QROWD appli­ca­tions can take advan­tage of the best of both worlds: the accu­ra­cy and scale of machine com­pu­ta­tion, and the skills, knowl­edge and exper­tise of peo­ple. We present the archi­tec­ture and main com­po­nents of the plat­form, as well as its usage to realise two mobil­i­ty use cas­es: esti­mat­ing the modal split, which refers to trips peo­ple take that involve more than one type of trans­port, and urban audit­ing.},
  address = {Cham},
  author = {Ib{á}{\~{n}}ez, Luis-Daniel and Mad­dale­na, Eddy and Gomer, Richard and Sim­perl, Ele­na and Zeni, Mat­tia and Big­not­ti, Enri­co and Chenu-Abente, Ronald and Giunchiglia, Faus­to and West­phal, Patrick and Stadler, Claus and Dzi­wis, Gor­dian and Lehmann, Jens and Yumusak, Semih and Voigt, Mar­tin and San­guino, Maria-Ange­les and Villaz{á}n, Javier and Ruiz, Ricar­do and Pari­ente-Lobo, Tomas},
  book­ti­tle = {Sus­tain­able Smart Cities: The­o­ret­i­cal Foun­da­tions and Prac­ti­cal Con­sid­er­a­tions},
  edi­tor = {Singh, Pradeep Kumar and Paprzy­c­ki, Marcin and Essaai­di, Mohamad and Rahi­mi, Shahram},
  key­words = {sys:relevantFor:infai},
  pages = {285–321},
  pub­lish­er = {Springer Inter­na­tion­al Pub­lish­ing},
  title = {QROWD—A Plat­form for Inte­grat­ing Cit­i­zens in Smart City Data Ana­lyt­ics},
  year = 2023
  }
  

  %0 Book Sec­tion
  %1 Ibanez2023-qrowd
  %A Ib{á}{\~{n}}ez, Luis-Daniel
  %A Mad­dale­na, Eddy
  %A Gomer, Richard
  %A Sim­perl, Ele­na
  %A Zeni, Mat­tia
  %A Big­not­ti, Enri­co
  %A Chenu-Abente, Ronald
  %A Giunchiglia, Faus­to
  %A West­phal, Patrick
  %A Stadler, Claus
  %A Dzi­wis, Gor­dian
  %A Lehmann, Jens
  %A Yumusak, Semih
  %A Voigt, Mar­tin
  %A San­guino, Maria-Ange­les
  %A Villaz{á}n, Javier
  %A Ruiz, Ricar­do
  %A Pari­ente-Lobo, Tomas
  %B Sus­tain­able Smart Cities: The­o­ret­i­cal Foun­da­tions and Prac­ti­cal Con­sid­er­a­tions
  %C Cham
  %D 2023
  %E Singh, Pradeep Kumar
  %E Paprzy­c­ki, Marcin
  %E Essaai­di, Mohamad
  %E Rahi­mi, Shahram
  %I Springer Inter­na­tion­al Pub­lish­ing
  %P 285–321
  %R 10.1007/978–3‑031–08815-5_16
  %T QROWD—A Plat­form for Inte­grat­ing Cit­i­zens in Smart City Data Ana­lyt­ics
  %U https://svn.aksw.org/papers/2022/SSC_qrowd/public.pdf
  %X Opti­miz­ing mobil­i­ty ser­vices is one of the great­est chal­lenges Smart Cities face in their efforts to improve res­i­dents’ well­be­ing and reduce {\$}{\$}{\backslash}text {\{}CO{\}}{\_}{\{}2{\}}{\$}{\$}emissions. The advent of IoT has cre­at­ed unpar­al­leled oppor­tu­ni­ties to col­lect large amounts of data about how peo­ple use trans­porta­tion. This data could be used to ascer­tain the qual­i­ty and reach of the ser­vices offered and to inform future policy—provided cities have the capa­bil­i­ties to process, curate, inte­grate and analyse the data effec­tive­ly. At the same time, to be tru­ly ‘Smart’, cities need to ensure that the data-dri­ven deci­sions they make reflect the needs of their cit­i­zens, cre­ate feed­back loops, and widen par­tic­i­pa­tion. In this chap­ter, we intro­duce QROWD, a data inte­gra­tion and ana­lyt­ics plat­form that seam­less­ly inte­grates mul­ti­ple data sources along­side human, social and com­pu­ta­tion­al intel­li­gence to build hybrid, auto­mat­ed data-cen­tric work­flows. By doing so, QROWD appli­ca­tions can take advan­tage of the best of both worlds: the accu­ra­cy and scale of machine com­pu­ta­tion, and the skills, knowl­edge and exper­tise of peo­ple. We present the archi­tec­ture and main com­po­nents of the plat­form, as well as its usage to realise two mobil­i­ty use cas­es: esti­mat­ing the modal split, which refers to trips peo­ple take that involve more than one type of trans­port, and urban audit­ing.
  %@ 978–3‑031–08815‑5
  

2022

• {LSQ} 2.0: A linked dataset of {SPARQL} query logsIn: Philippe Cud­ré-Mau­roux (ed.) Seman­tic Web, IOS Press, pp. 1–23Claus Stadler, Muham­mad Saleem, Qais­er Mehmood, Car­los Buil-Aran­da, Michel Dumon­tier, Aidan Hogan and Axel-Cyrille Ngonga Ngomo

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  We present the Linked SPARQL Queries (LSQ) dataset, which cur­rent­ly describes 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries extract­ed from the logs of 27 dif­fer­ent end­points. The LSQ dataset pro­vides RDF descrip­tions of each such query, which are indexed in a pub­lic LSQ end­point, allow­ing inter­est­ed par­ties to find queries with the char­ac­ter­is­tics they require. We begin by describ­ing the use cas­es envis­aged for the LSQ dataset, which include appli­ca­tions for research on com­mon fea­tures of queries, for build­ing cus­tom bench­marks, and for design­ing user inter­faces. We then dis­cuss how LSQ has been used in prac­tice since the release of four ini­tial SPARQL logs in 2015. We dis­cuss the mod­el and vocab­u­lary that we use to rep­re­sent these queries in RDF. We then pro­vide a brief overview of the 27 end­points from which we extract­ed queries in terms of the domain to which they per­tain and the data they con­tain. We pro­vide sta­tis­tics on the queries includ­ed from each log, includ­ing the num­ber of query exe­cu­tions, unique queries, as well as dis­tri­b­u­tions of queries for a vari­ety of select­ed char­ac­ter­is­tics. We final­ly dis­cuss how the LSQ dataset is host­ed and how it can be accessed and lever­aged by inter­est­ed par­ties for their use cases.

  @article{stadler2022-lsq20,
  abstract = {We present the Linked SPARQL Queries (LSQ) dataset, which cur­rent­ly describes 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries extract­ed from the logs of 27 dif­fer­ent end­points. The LSQ dataset pro­vides RDF descrip­tions of each such query, which are indexed in a pub­lic LSQ end­point, allow­ing inter­est­ed par­ties to find queries with the char­ac­ter­is­tics they require. We begin by describ­ing the use cas­es envis­aged for the LSQ dataset, which include appli­ca­tions for research on com­mon fea­tures of queries, for build­ing cus­tom bench­marks, and for design­ing user inter­faces. We then dis­cuss how LSQ has been used in prac­tice since the release of four ini­tial SPARQL logs in 2015. We dis­cuss the mod­el and vocab­u­lary that we use to rep­re­sent these queries in RDF. We then pro­vide a brief overview of the 27 end­points from which we extract­ed queries in terms of the domain to which they per­tain and the data they con­tain. We pro­vide sta­tis­tics on the queries includ­ed from each log, includ­ing the num­ber of query exe­cu­tions, unique queries, as well as dis­tri­b­u­tions of queries for a vari­ety of select­ed char­ac­ter­is­tics. We final­ly dis­cuss how the LSQ dataset is host­ed and how it can be accessed and lever­aged by inter­est­ed par­ties for their use cas­es.},
  author = {Stadler, Claus and Saleem, Muham­mad and Mehmood, Qais­er and Buil-Aran­da, Car­los and Dumon­tier, Michel and Hogan, Aidan and Ngonga Ngo­mo, Axel-Cyrille},
  edi­tor = {Cud­ré-Mau­roux, Philippe},
  jour­nal = {Seman­tic Web},
  key­words = {sys:relevantFor:infai},
  month = 11,
  pages = {1–23},
  pub­lish­er = {IOS Press},
  title = {{LSQ} 2.0: A linked dataset of {SPARQL} query logs},
  year = 2022
  }
  

  %0 Jour­nal Arti­cle
  %1 stadler2022-lsq20
  %A Stadler, Claus
  %A Saleem, Muham­mad
  %A Mehmood, Qais­er
  %A Buil-Aran­da, Car­los
  %A Dumon­tier, Michel
  %A Hogan, Aidan
  %A Ngonga Ngo­mo, Axel-Cyrille
  %D 2022
  %E Cud­ré-Mau­roux, Philippe
  %I IOS Press
  %J Seman­tic Web
  %P 1–23
  %R 10.3233/SW-223015
  %T {LSQ} 2.0: A linked dataset of {SPARQL} query logs
  %U https://www.semantic-web-journal.net/system/files/swj3015.pdf
  %X We present the Linked SPARQL Queries (LSQ) dataset, which cur­rent­ly describes 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries extract­ed from the logs of 27 dif­fer­ent end­points. The LSQ dataset pro­vides RDF descrip­tions of each such query, which are indexed in a pub­lic LSQ end­point, allow­ing inter­est­ed par­ties to find queries with the char­ac­ter­is­tics they require. We begin by describ­ing the use cas­es envis­aged for the LSQ dataset, which include appli­ca­tions for research on com­mon fea­tures of queries, for build­ing cus­tom bench­marks, and for design­ing user inter­faces. We then dis­cuss how LSQ has been used in prac­tice since the release of four ini­tial SPARQL logs in 2015. We dis­cuss the mod­el and vocab­u­lary that we use to rep­re­sent these queries in RDF. We then pro­vide a brief overview of the 27 end­points from which we extract­ed queries in terms of the domain to which they per­tain and the data they con­tain. We pro­vide sta­tis­tics on the queries includ­ed from each log, includ­ing the num­ber of query exe­cu­tions, unique queries, as well as dis­tri­b­u­tions of queries for a vari­ety of select­ed char­ac­ter­is­tics. We final­ly dis­cuss how the LSQ dataset is host­ed and how it can be accessed and lever­aged by inter­est­ed par­ties for their use cases.
  

• Seman­tifi­ca­tion of Geospa­tial Infor­ma­tion for Enriched Knowl­edge Rep­re­sen­ta­tion in Con­text of Cri­sis Infor­mat­icsIn: Natanael Arndt, Sabine Grün­der-Fahrer, Julia Holze, Michael Mar­tin and Sebas­t­ian Tramp (eds.): Pro­ceed­ings of the Inter­na­tion­al Work­shop on Data-dri­ven Resilience Research 2022, {CEUR} Work­shop Pro­ceed­ings. vol. 3376. Leipzig, Ger­manyClaus Stadler, Simon Bin, Lorenz Büh­mann, Nor­man Radtke, Kurt Jung­hanns, Sabine Grün­der-Fahrer and Michael Martin

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  In the con­text of cri­sis infor­mat­ics, the inte­gra­tion and exploita­tion of high vol­umes of het­ero­ge­neous data from mul­ti­ple sources is one of the big chances as well as chal­lenges up to now. Seman­tic Web tech­nolo­gies have proven a valu­able means to inte­grate and rep­re­sent knowl­edge on the basis of domain con­cepts which improves inter­op­er­abil­i­ty and inter­pretabil­i­ty of infor­ma­tion resources and allows deriv­ing more knowl­edge via seman­tic rela­tions and rea­son­ing. In this paper, we inves­ti­gate the poten­tial of rep­re­sent­ing and pro­cess­ing geospa­tial infor­ma­tion with­in the seman­tic par­a­digm. We show, on the tech­ni­cal lev­el, how exist­ing open source means can be used and sup­ple­ment­ed as to effi­cient­ly han­dle geo­graph­ic infor­ma­tion and to con­vey exem­plary results high­ly rel­e­vant in con­text of cri­sis man­age­ment appli­ca­tions. When giv­en seman­tic resources get enriched with geospa­tial infor­ma­tion, new infor­ma­tion can be retrieved com­bin­ing the con­cepts of mul­ti-poly­gons and geo-coor­di­nates and using the pow­er of GeoSPAR­QL queries. Cus­tom SPARQL exten­sion func­tions and data types for JSON, XML and CSV as well as for dialects such as Geo­J­SON and GML allow for suc­cinct inte­gra­tion of het­ero­ge­neous data. We imple­ment­ed these fea­tures for the Apache Jena Seman­tic Web frame­work by lever­ag­ing its plu­g­in sys­tems. Fur­ther­more, sig­nif­i­cant improve­ments w.r.t. GeoSPAR­QL query per­for­mance have been con­tributed to the framework.

  @inproceedings{stadler-c-2022–geospacial,
  abstract = {In the con­text of cri­sis infor­mat­ics, the inte­gra­tion and exploita­tion of high vol­umes of het­ero­ge­neous data from mul­ti­ple sources is one of the big chances as well as chal­lenges up to now. Seman­tic Web tech­nolo­gies have proven a valu­able means to inte­grate and rep­re­sent knowl­edge on the basis of domain con­cepts which improves inter­op­er­abil­i­ty and inter­pretabil­i­ty of infor­ma­tion resources and allows deriv­ing more knowl­edge via seman­tic rela­tions and rea­son­ing. In this paper, we inves­ti­gate the poten­tial of rep­re­sent­ing and pro­cess­ing geospa­tial infor­ma­tion with­in the seman­tic par­a­digm. We show, on the tech­ni­cal lev­el, how exist­ing open source means can be used and sup­ple­ment­ed as to effi­cient­ly han­dle geo­graph­ic infor­ma­tion and to con­vey exem­plary results high­ly rel­e­vant in con­text of cri­sis man­age­ment appli­ca­tions. When giv­en seman­tic resources get enriched with geospa­tial infor­ma­tion, new infor­ma­tion can be retrieved com­bin­ing the con­cepts of mul­ti-poly­gons and geo-coor­di­nates and using the pow­er of GeoSPAR­QL queries. Cus­tom SPARQL exten­sion func­tions and data types for JSON, XML and CSV as well as for dialects such as Geo­J­SON and GML allow for suc­cinct inte­gra­tion of het­ero­ge­neous data. We imple­ment­ed these fea­tures for the Apache Jena Seman­tic Web frame­work by lever­ag­ing its plu­g­in sys­tems. Fur­ther­more, sig­nif­i­cant improve­ments w.r.t. GeoSPAR­QL query per­for­mance have been con­tributed to the frame­work.},
  address = {Leipzig, Ger­many},
  author = {Stadler, Claus and Bin, Simon and Büh­mann, Lorenz and Radtke, Nor­man and Jung­hanns, Kurt and Grün­der-Fahrer, Sabine and Mar­tin, Michael},
  book­ti­tle = {Pro­ceed­ings of the Inter­na­tion­al Work­shop on Data-dri­ven Resilience Research 2022},
  edi­tor = {Arndt, Natanael and Grün­der-Fahrer, Sabine and Holze, Julia and Mar­tin, Michael and Tramp, Sebas­t­ian},
  key­words = {sys:relevantFor:infai},
  month = {07},
  series = {{CEUR} Work­shop Pro­ceed­ings},
  title = {Seman­tifi­ca­tion of Geospa­tial Infor­ma­tion for Enriched Knowl­edge Rep­re­sen­ta­tion in Con­text of Cri­sis Infor­mat­ics},
  vol­ume = 3376,
  year = 2022
  }
  

  %0 Con­fer­ence Paper
  %1 stadler-c-2022–geospacial
  %A Stadler, Claus
  %A Bin, Simon
  %A Büh­mann, Lorenz
  %A Radtke, Nor­man
  %A Jung­hanns, Kurt
  %A Grün­der-Fahrer, Sabine
  %A Mar­tin, Michael
  %B Pro­ceed­ings of the Inter­na­tion­al Work­shop on Data-dri­ven Resilience Research 2022
  %C Leipzig, Ger­many
  %D 2022
  %E Arndt, Natanael
  %E Grün­der-Fahrer, Sabine
  %E Holze, Julia
  %E Mar­tin, Michael
  %E Tramp, Sebas­t­ian
  %T Seman­tifi­ca­tion of Geospa­tial Infor­ma­tion for Enriched Knowl­edge Rep­re­sen­ta­tion in Con­text of Cri­sis Infor­mat­ics
  %U https://ceur-ws.org/Vol-3376/paper03.pdf
  %V 3376
  %X In the con­text of cri­sis infor­mat­ics, the inte­gra­tion and exploita­tion of high vol­umes of het­ero­ge­neous data from mul­ti­ple sources is one of the big chances as well as chal­lenges up to now. Seman­tic Web tech­nolo­gies have proven a valu­able means to inte­grate and rep­re­sent knowl­edge on the basis of domain con­cepts which improves inter­op­er­abil­i­ty and inter­pretabil­i­ty of infor­ma­tion resources and allows deriv­ing more knowl­edge via seman­tic rela­tions and rea­son­ing. In this paper, we inves­ti­gate the poten­tial of rep­re­sent­ing and pro­cess­ing geospa­tial infor­ma­tion with­in the seman­tic par­a­digm. We show, on the tech­ni­cal lev­el, how exist­ing open source means can be used and sup­ple­ment­ed as to effi­cient­ly han­dle geo­graph­ic infor­ma­tion and to con­vey exem­plary results high­ly rel­e­vant in con­text of cri­sis man­age­ment appli­ca­tions. When giv­en seman­tic resources get enriched with geospa­tial infor­ma­tion, new infor­ma­tion can be retrieved com­bin­ing the con­cepts of mul­ti-poly­gons and geo-coor­di­nates and using the pow­er of GeoSPAR­QL queries. Cus­tom SPARQL exten­sion func­tions and data types for JSON, XML and CSV as well as for dialects such as Geo­J­SON and GML allow for suc­cinct inte­gra­tion of het­ero­ge­neous data. We imple­ment­ed these fea­tures for the Apache Jena Seman­tic Web frame­work by lever­ag­ing its plu­g­in sys­tems. Fur­ther­more, sig­nif­i­cant improve­ments w.r.t. GeoSPAR­QL query per­for­mance have been con­tributed to the framework.
  

• {LSQ} Frame­work: The {LSQ} Frame­work for {SPARQL} Query Log Pro­cess­ingIn: 6th Work­shop on Stor­ing, Query­ing and Bench­mark­ing Knowl­edge Graphs @ ISWC 2022, {CEUR} Work­shop Pro­ceed­ings. vol. 3279Claus Stadler, Muham­mad Saleem and Axel-Cyrille Ngonga Ngomo

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  The Linked SPARQL Queries (LSQ) datasets con­tain real-world SPARQL queries col­lect­ed from the query logs of the pub­licly avail­able SPARQL end­points. In LSQ, each SPARQL query is rep­re­sent­ed as RDF with var­i­ous struc­tur­al and data-dri­ven fea­tures attached. In this paper, we present the LSQ Java frame­work for cre­at­ing rich knowl­edge graphs from SPARQL query logs. The frame­work is able to RDFize SPARQL query logs, which are avail­able in dif­fer­ent for­mats, in a scal­able way. Fur­ther­more, the frame­work offers a set of sta­t­ic and dynam­ic enrich­ers. Sta­t­ic enrich­ers derive infor­ma­tion from the queries, such as their num­ber of basic graph pat­terns and pro­ject­ed vari­ables or even a full SPIN mod­el. Dynam­ic enrich­ment involves addi­tion­al resources. For instance, the bench­mark enrich­er exe­cutes queries against a SPARQL end­point and col­lects query exe­cu­tion times and result set sizes. This frame­work has already been used to con­vert query logs of 27 pub­lic SPARQL end­points, rep­re­sent­ing 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries. The LSQ queries have been used in many use cas­es such as bench­mark­ing based on real-world SPARQL queries, SPARQL adop­tion, caching, query opti­miza­tion, use­abil­i­ty analy­sis, and meta-query­ing. Real­iza­tion of LSQ required devis­ing nov­el soft­ware com­po­nents to (a) improve scal­a­bil­i­ty of RDF data pro­cess­ing with the Apache Spark Big Data frame­work and (b) ease oper­a­tions of com­plex RDF data mod­els such as con­trolled skolem­iza­tion. Fol­low­ing the spir­it of Open­Source soft­ware devel­op­ment and the “don’t repeat your­self” (DRY) par­a­digm, the work on the LSQ frame­work also result­ed in con­tri­bu­tions to Apache Jena in order to make these improve­ments read­i­ly avail­able out­side of the LSQ context.

  @inproceedings{stadler2022-lsq-framework,
  abstract = {The Linked SPARQL Queries (LSQ) datasets con­tain real-world SPARQL queries col­lect­ed from the query logs of the pub­licly avail­able SPARQL end­points. In LSQ, each SPARQL query is rep­re­sent­ed as RDF with var­i­ous struc­tur­al and data-dri­ven fea­tures attached. In this paper, we present the LSQ Java frame­work for cre­at­ing rich knowl­edge graphs from SPARQL query logs. The frame­work is able to RDFize SPARQL query logs, which are avail­able in dif­fer­ent for­mats, in a scal­able way. Fur­ther­more, the frame­work offers a set of sta­t­ic and dynam­ic enrich­ers. Sta­t­ic enrich­ers derive infor­ma­tion from the queries, such as their num­ber of basic graph pat­terns and pro­ject­ed vari­ables or even a full SPIN mod­el. Dynam­ic enrich­ment involves addi­tion­al resources. For instance, the bench­mark enrich­er exe­cutes queries against a SPARQL end­point and col­lects query exe­cu­tion times and result set sizes. This frame­work has already been used to con­vert query logs of 27 pub­lic SPARQL end­points, rep­re­sent­ing 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries. The LSQ queries have been used in many use cas­es such as bench­mark­ing based on real-world SPARQL queries, SPARQL adop­tion, caching, query opti­miza­tion, use­abil­i­ty analy­sis, and meta-query­ing. Real­iza­tion of LSQ required devis­ing nov­el soft­ware com­po­nents to (a) improve scal­a­bil­i­ty of RDF data pro­cess­ing with the Apache Spark Big Data frame­work and (b) ease oper­a­tions of com­plex RDF data mod­els such as con­trolled skolem­iza­tion. Fol­low­ing the spir­it of Open­Source soft­ware devel­op­ment and the “don’t repeat your­self” (DRY) par­a­digm, the work on the LSQ frame­work also result­ed in con­tri­bu­tions to Apache Jena in order to make these improve­ments read­i­ly avail­able out­side of the LSQ con­text.},
  author = {Stadler, Claus and Saleem, Muham­mad and Ngo­mo, Axel-Cyrille Ngonga},
  book­ti­tle = {6th Work­shop on Stor­ing, Query­ing and Bench­mark­ing Knowl­edge Graphs @ ISWC 2022},
  key­words = {sys:relevantFor:infai},
  series = {{CEUR} Work­shop Pro­ceed­ings},
  title = {{LSQ} Frame­work: The {LSQ} Frame­work for {SPARQL} Query Log Pro­cess­ing},
  vol­ume = 3279,
  year = 2022
  }
  

  %0 Con­fer­ence Paper
  %1 stadler2022-lsq-frame­work
  %A Stadler, Claus
  %A Saleem, Muham­mad
  %A Ngo­mo, Axel-Cyrille Ngonga
  %B 6th Work­shop on Stor­ing, Query­ing and Bench­mark­ing Knowl­edge Graphs @ ISWC 2022
  %D 2022
  %T {LSQ} Frame­work: The {LSQ} Frame­work for {SPARQL} Query Log Pro­cess­ing
  %U https://ceur-ws.org/Vol-3279/paper4.pdf
  %V 3279
  %X The Linked SPARQL Queries (LSQ) datasets con­tain real-world SPARQL queries col­lect­ed from the query logs of the pub­licly avail­able SPARQL end­points. In LSQ, each SPARQL query is rep­re­sent­ed as RDF with var­i­ous struc­tur­al and data-dri­ven fea­tures attached. In this paper, we present the LSQ Java frame­work for cre­at­ing rich knowl­edge graphs from SPARQL query logs. The frame­work is able to RDFize SPARQL query logs, which are avail­able in dif­fer­ent for­mats, in a scal­able way. Fur­ther­more, the frame­work offers a set of sta­t­ic and dynam­ic enrich­ers. Sta­t­ic enrich­ers derive infor­ma­tion from the queries, such as their num­ber of basic graph pat­terns and pro­ject­ed vari­ables or even a full SPIN mod­el. Dynam­ic enrich­ment involves addi­tion­al resources. For instance, the bench­mark enrich­er exe­cutes queries against a SPARQL end­point and col­lects query exe­cu­tion times and result set sizes. This frame­work has already been used to con­vert query logs of 27 pub­lic SPARQL end­points, rep­re­sent­ing 43.95 mil­lion exe­cu­tions of 11.56 mil­lion unique SPARQL queries. The LSQ queries have been used in many use cas­es such as bench­mark­ing based on real-world SPARQL queries, SPARQL adop­tion, caching, query opti­miza­tion, use­abil­i­ty analy­sis, and meta-query­ing. Real­iza­tion of LSQ required devis­ing nov­el soft­ware com­po­nents to (a) improve scal­a­bil­i­ty of RDF data pro­cess­ing with the Apache Spark Big Data frame­work and (b) ease oper­a­tions of com­plex RDF data mod­els such as con­trolled skolem­iza­tion. Fol­low­ing the spir­it of Open­Source soft­ware devel­op­ment and the “don’t repeat your­self” (DRY) par­a­digm, the work on the LSQ frame­work also result­ed in con­tri­bu­tions to Apache Jena in order to make these improve­ments read­i­ly avail­able out­side of the LSQ context.
  

2021

• Towards the next gen­er­a­tion of the Linked­Ge­o­Da­ta project using vir­tu­al knowl­edge graphsIn: Jour­nal of Web Seman­tics vol. 71, p. 100662Lin­fang Ding, Guo­hui Xiao, Albulen Pano, Claus Stadler and Diego Calvanese

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  With the advance­ment of Seman­tic Tech­nolo­gies, large geospa­tial data sources have been increas­ing­ly pub­lished as Linked data on the Web. The Linked­Ge­o­Da­ta project is one of the most promi­nent such projects to cre­ate a large knowl­edge graph from Open­StreetMap (OSM) with glob­al cov­er­age and inter­link­ing of oth­er data sources. In this paper, we report on the ongo­ing effort of expos­ing the rela­tion­al data­base in Linked­Ge­o­Da­ta as a SPARQL end­point using Vir­tu­al Knowl­edge Graph (VKG) tech­nol­o­gy. Specif­i­cal­ly, we present two real­iza­tions of VKGs, using the two sys­tems Spar­qli­fy and Ontop. In order to improve com­pli­ance with the OGC GeoSPAR­QL stan­dard, we have imple­ment­ed GeoSPAR­QL sup­port in Ontop v4. More­over, we have eval­u­at­ed the VKG-pow­ered Linked­Ge­o­Da­ta in the test areas of Italy and Ger­many. Our exper­i­ments demon­strate that such sys­tem sup­ports com­plex GeoSPAR­QL queries, which con­firms that query answer­ing in the VKG approach is efficient.

  @article{Ding2021,
  abstract = {With the advance­ment of Seman­tic Tech­nolo­gies, large geospa­tial data sources have been increas­ing­ly pub­lished as Linked data on the Web. The Linked­Ge­o­Da­ta project is one of the most promi­nent such projects to cre­ate a large knowl­edge graph from Open­StreetMap (OSM) with glob­al cov­er­age and inter­link­ing of oth­er data sources. In this paper, we report on the ongo­ing effort of expos­ing the rela­tion­al data­base in Linked­Ge­o­Da­ta as a SPARQL end­point using Vir­tu­al Knowl­edge Graph (VKG) tech­nol­o­gy. Specif­i­cal­ly, we present two real­iza­tions of VKGs, using the two sys­tems Spar­qli­fy and Ontop. In order to improve com­pli­ance with the OGC GeoSPAR­QL stan­dard, we have imple­ment­ed GeoSPAR­QL sup­port in Ontop v4. More­over, we have eval­u­at­ed the VKG-pow­ered Linked­Ge­o­Da­ta in the test areas of Italy and Ger­many. Our exper­i­ments demon­strate that such sys­tem sup­ports com­plex GeoSPAR­QL queries, which con­firms that query answer­ing in the VKG approach is effi­cient.},
  author = {Ding, Lin­fang and Xiao, Guo­hui and Pano, Albulen and Stadler, Claus and Cal­vanese, Diego},
  jour­nal = {Jour­nal of Web Seman­tics},
  key­words = {sys:relevantFor:infai},
  pages = 100662,
  title = {Towards the next gen­er­a­tion of the Linked­Ge­o­Da­ta project using vir­tu­al knowl­edge graphs},
  vol­ume = 71,
  year = 2021
  }
  

  %0 Jour­nal Arti­cle
  %1 Ding2021
  %A Ding, Lin­fang
  %A Xiao, Guo­hui
  %A Pano, Albulen
  %A Stadler, Claus
  %A Cal­vanese, Diego
  %D 2021
  %J Jour­nal of Web Seman­tics
  %P 100662
  %R https://doi.org/10.1016/j.websem.2021.100662
  %T Towards the next gen­er­a­tion of the Linked­Ge­o­Da­ta project using vir­tu­al knowl­edge graphs
  %U https://svn.aksw.org/papers/2023/JWS_LinkedGeoData_Ontop/public.pdf
  %V 71
  %X With the advance­ment of Seman­tic Tech­nolo­gies, large geospa­tial data sources have been increas­ing­ly pub­lished as Linked data on the Web. The Linked­Ge­o­Da­ta project is one of the most promi­nent such projects to cre­ate a large knowl­edge graph from Open­StreetMap (OSM) with glob­al cov­er­age and inter­link­ing of oth­er data sources. In this paper, we report on the ongo­ing effort of expos­ing the rela­tion­al data­base in Linked­Ge­o­Da­ta as a SPARQL end­point using Vir­tu­al Knowl­edge Graph (VKG) tech­nol­o­gy. Specif­i­cal­ly, we present two real­iza­tions of VKGs, using the two sys­tems Spar­qli­fy and Ontop. In order to improve com­pli­ance with the OGC GeoSPAR­QL stan­dard, we have imple­ment­ed GeoSPAR­QL sup­port in Ontop v4. More­over, we have eval­u­at­ed the VKG-pow­ered Linked­Ge­o­Da­ta in the test areas of Italy and Ger­many. Our exper­i­ments demon­strate that such sys­tem sup­ports com­plex GeoSPAR­QL queries, which con­firms that query answer­ing in the VKG approach is efficient.
  

• {DistRDF2ML} — Scal­able Dis­trib­uted In-Mem­o­ry Machine Learn­ing Pipelines for {RDF} Knowl­edge GraphsIn: Pro­ceed­ings of the 30th ACM Inter­na­tion­al Con­fer­ence on Infor­ma­tion \& Knowl­edge Man­age­ment, CIKM ’21. Vir­tu­al Event, Queens­land, Aus­tralia : Asso­ci­a­tion for Com­put­ing Machin­ery — ISBN 9781450384469, pp. 4465–4474Carsten Felix Draschn­er, Claus Stadler, Far­shad Bakhshan­de­gan Moghad­dam, Jens Lehmann and Haji­ra Jabeen

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  This paper presents DistRDF2ML, the gener­ic, scal­able, and dis­trib­uted frame­work for cre­at­ing in-mem­o­ry data pre­pro­cess­ing pipelines for Spark-based machine learn­ing on RDF knowl­edge graphs. This frame­work intro­duces soft­ware mod­ules that trans­form large-scale RDF data into ML-ready fixed-length numer­ic fea­ture vec­tors. The devel­oped mod­ules are opti­mized to the mul­ti-modal nature of knowl­edge graphs. DistRDF2ML pro­vides aligned soft­ware design and usage prin­ci­ples as com­mon data sci­ence stacks that offer an easy-to-use pack­age for cre­at­ing machine learn­ing pipelines. The mod­ules used in the pipeline, the hyper-para­me­ters and the results are export­ed as a seman­tic struc­ture that can be used to enrich the orig­i­nal knowl­edge graph. The seman­tic rep­re­sen­ta­tion of meta­da­ta and machine learn­ing results offers the advan­tage of increas­ing the machine learn­ing pipelines’ reusabil­i­ty, explain­abil­i­ty, and repro­ducibil­i­ty. The entire frame­work of DistRDF2ML is open source, inte­grat­ed into the holis­tic SANSA stack, doc­u­ment­ed in scala-docs, and cov­ered by unit tests. DistRDF2ML demon­strates its scal­able design across dif­fer­ent pro­cess­ing pow­er con­fig­u­ra­tions and (hyper-)parameter setups with­in var­i­ous exper­i­ments. The frame­work brings the three worlds of knowl­edge graph engi­neers, dis­trib­uted com­pu­ta­tion devel­op­ers, and data sci­en­tists clos­er togeth­er and offers all of them the cre­ation of explain­able ML pipelines using a few lines of code.

  @inproceedings{Draschner2021,
  abstract = {This paper presents DistRDF2ML, the gener­ic, scal­able, and dis­trib­uted frame­work for cre­at­ing in-mem­o­ry data pre­pro­cess­ing pipelines for Spark-based machine learn­ing on RDF knowl­edge graphs. This frame­work intro­duces soft­ware mod­ules that trans­form large-scale RDF data into ML-ready fixed-length numer­ic fea­ture vec­tors. The devel­oped mod­ules are opti­mized to the mul­ti-modal nature of knowl­edge graphs. DistRDF2ML pro­vides aligned soft­ware design and usage prin­ci­ples as com­mon data sci­ence stacks that offer an easy-to-use pack­age for cre­at­ing machine learn­ing pipelines. The mod­ules used in the pipeline, the hyper-para­me­ters and the results are export­ed as a seman­tic struc­ture that can be used to enrich the orig­i­nal knowl­edge graph. The seman­tic rep­re­sen­ta­tion of meta­da­ta and machine learn­ing results offers the advan­tage of increas­ing the machine learn­ing pipelines’ reusabil­i­ty, explain­abil­i­ty, and repro­ducibil­i­ty. The entire frame­work of DistRDF2ML is open source, inte­grat­ed into the holis­tic SANSA stack, doc­u­ment­ed in scala-docs, and cov­ered by unit tests. DistRDF2ML demon­strates its scal­able design across dif­fer­ent pro­cess­ing pow­er con­fig­u­ra­tions and (hyper-)parameter setups with­in var­i­ous exper­i­ments. The frame­work brings the three worlds of knowl­edge graph engi­neers, dis­trib­uted com­pu­ta­tion devel­op­ers, and data sci­en­tists clos­er togeth­er and offers all of them the cre­ation of explain­able ML pipelines using a few lines of code.},
  address = {New York, NY, USA},
  author = {Draschn­er, Carsten Felix and Stadler, Claus and Bakhshan­de­gan Moghad­dam, Far­shad and Lehmann, Jens and Jabeen, Haji­ra},
  book­ti­tle = {Pro­ceed­ings of the 30th ACM Inter­na­tion­al Con­fer­ence on Infor­ma­tion \& Knowl­edge Man­age­ment},
  key­words = {sys:relevantFor:infai},
  pages = {4465–4474},
  pub­lish­er = {Asso­ci­a­tion for Com­put­ing Machin­ery},
  series = {CIKM ’21},
  title = {{DistRDF2ML} — Scal­able Dis­trib­uted In-Mem­o­ry Machine Learn­ing Pipelines for {RDF} Knowl­edge Graphs},
  year = 2021
  }
  

  %0 Con­fer­ence Paper
  %1 Draschner2021
  %A Draschn­er, Carsten Felix
  %A Stadler, Claus
  %A Bakhshan­de­gan Moghad­dam, Far­shad
  %A Lehmann, Jens
  %A Jabeen, Haji­ra
  %B Pro­ceed­ings of the 30th ACM Inter­na­tion­al Con­fer­ence on Infor­ma­tion \& Knowl­edge Man­age­ment
  %C New York, NY, USA
  %D 2021
  %I Asso­ci­a­tion for Com­put­ing Machin­ery
  %P 4465–4474
  %R 10.1145/3459637.3481999
  %T {DistRDF2ML} — Scal­able Dis­trib­uted In-Mem­o­ry Machine Learn­ing Pipelines for {RDF} Knowl­edge Graphs
  %U https://svn.aksw.org/papers/2021/cikm-distrdf2ml/public.pdf
  %X This paper presents DistRDF2ML, the gener­ic, scal­able, and dis­trib­uted frame­work for cre­at­ing in-mem­o­ry data pre­pro­cess­ing pipelines for Spark-based machine learn­ing on RDF knowl­edge graphs. This frame­work intro­duces soft­ware mod­ules that trans­form large-scale RDF data into ML-ready fixed-length numer­ic fea­ture vec­tors. The devel­oped mod­ules are opti­mized to the mul­ti-modal nature of knowl­edge graphs. DistRDF2ML pro­vides aligned soft­ware design and usage prin­ci­ples as com­mon data sci­ence stacks that offer an easy-to-use pack­age for cre­at­ing machine learn­ing pipelines. The mod­ules used in the pipeline, the hyper-para­me­ters and the results are export­ed as a seman­tic struc­ture that can be used to enrich the orig­i­nal knowl­edge graph. The seman­tic rep­re­sen­ta­tion of meta­da­ta and machine learn­ing results offers the advan­tage of increas­ing the machine learn­ing pipelines’ reusabil­i­ty, explain­abil­i­ty, and repro­ducibil­i­ty. The entire frame­work of DistRDF2ML is open source, inte­grat­ed into the holis­tic SANSA stack, doc­u­ment­ed in scala-docs, and cov­ered by unit tests. DistRDF2ML demon­strates its scal­able design across dif­fer­ent pro­cess­ing pow­er con­fig­u­ra­tions and (hyper-)parameter setups with­in var­i­ous exper­i­ments. The frame­work brings the three worlds of knowl­edge graph engi­neers, dis­trib­uted com­pu­ta­tion devel­op­ers, and data sci­en­tists clos­er togeth­er and offers all of them the cre­ation of explain­able ML pipelines using a few lines of code.
  %@ 9781450384469
  

• Open Data and the Sta­tus Quo — A Fine-Grained Eval­u­a­tion Frame­work for Open Data Qual­i­ty and an Analy­sis of Open Data por­tals in Ger­manyIn: Lisa Wenige, Claus Stadler, Michael Mar­tin, Richard Figu­ra, Robert Sauter and Christo­pher W. Frank

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{wenige2021open,
  author = {Wenige, Lisa and Stadler, Claus and Mar­tin, Michael and Figu­ra, Richard and Sauter, Robert and Frank, Christo­pher W.},
  key­words = {sys:relevantFor:infai},
  title = {Open Data and the Sta­tus Quo — A Fine-Grained Eval­u­a­tion Frame­work for Open Data Qual­i­ty and an Analy­sis of Open Data por­tals in Ger­many},
  year = 2021
  }
  

  %0 Con­fer­ence Paper
  %1 wenige2021open
  %A Wenige, Lisa
  %A Stadler, Claus
  %A Mar­tin, Michael
  %A Figu­ra, Richard
  %A Sauter, Robert
  %A Frank, Christo­pher W.
  %D 2021
  %T Open Data and the Sta­tus Quo — A Fine-Grained Eval­u­a­tion Frame­work for Open Data Qual­i­ty and an Analy­sis of Open Data por­tals in Ger­many
  %U https://arxiv.org/pdf/2106.09590.pdf
  

2020

• MINDS: A Trans­la­tor to Embed Math­e­mat­i­cal Expres­sions Inside SPARQL QueriesIn: Eva Blomqvist, Paul Groth, Vic­tor de Boer, Tas­si­lo Pel­le­gri­ni, Mehwish Alam, Tobias K{ä}fer, Peter Kiese­berg, Sab­ri­na Kir­rane, Albert Mero{\~{n}}o‑Pe{\~{n}}uela, et al. (eds.): Seman­tic Sys­tems. In the Era of Knowl­edge Graphs. Cham : Springer Inter­na­tion­al Pub­lish­ing — ISBN 978–3‑030–59833‑4, pp. 104–117Damien Graux, Gez­im Sejdiu, Claus Stadler, Giulio Napoli­tano and Jens Lehmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  The recent deploy­ments of seman­tic web tools and the expan­sion of avail­able linked datasets have giv­en users the oppor­tu­ni­ty of build­ing increas­ing­ly com­plex appli­ca­tions. These emerg­ing use cas­es often require queries con­tain­ing math­e­mat­i­cal for­mu­las such as euclid­ean dis­tances or unit con­ver­sions. Cur­rent­ly, the lat­est SPARQL stan­dard (ver­sion 1.1) only embeds basic math oper­a­tors. Thus, to address this short­com­ing, some pop­u­lar SPARQL eval­u­a­tors pro­vide built-in tools to cov­er spe­cif­ic needs; how­ev­er, such tools are not stan­dard yet. To offer users a more gener­ic solu­tion, we pro­pose and share MINDS, a trans­la­tor of math­e­mat­i­cal expres­sions into SPAR­QL-com­pli­ant bind­ings which can be under­stood by any eval­u­a­tor. MINDS there­by facil­i­tates the query design when­ev­er math­e­mat­i­cal com­pu­ta­tions are need­ed in a SPARQL query.

  @inproceedings{graux2020-minds,
  abstract = {The recent deploy­ments of seman­tic web tools and the expan­sion of avail­able linked datasets have giv­en users the oppor­tu­ni­ty of build­ing increas­ing­ly com­plex appli­ca­tions. These emerg­ing use cas­es often require queries con­tain­ing math­e­mat­i­cal for­mu­las such as euclid­ean dis­tances or unit con­ver­sions. Cur­rent­ly, the lat­est SPARQL stan­dard (ver­sion 1.1) only embeds basic math oper­a­tors. Thus, to address this short­com­ing, some pop­u­lar SPARQL eval­u­a­tors pro­vide built-in tools to cov­er spe­cif­ic needs; how­ev­er, such tools are not stan­dard yet. To offer users a more gener­ic solu­tion, we pro­pose and share MINDS, a trans­la­tor of math­e­mat­i­cal expres­sions into SPAR­QL-com­pli­ant bind­ings which can be under­stood by any eval­u­a­tor. MINDS there­by facil­i­tates the query design when­ev­er math­e­mat­i­cal com­pu­ta­tions are need­ed in a SPARQL query.},
  address = {Cham},
  author = {Graux, Damien and Sejdiu, Gez­im and Stadler, Claus and Napoli­tano, Giulio and Lehmann, Jens},
  book­ti­tle = {Seman­tic Sys­tems. In the Era of Knowl­edge Graphs},
  edi­tor = {Blomqvist, Eva and Groth, Paul and de Boer, Vic­tor and Pel­le­gri­ni, Tas­si­lo and Alam, Mehwish and K{ä}fer, Tobias and Kiese­berg, Peter and Kir­rane, Sab­ri­na and Mero{\~{n}}o‑Pe{\~{n}}uela, Albert and Pan­dit, Harsh­vard­han J.},
  key­words = {sys:relevantFor:infai},
  pages = {104–117},
  pub­lish­er = {Springer Inter­na­tion­al Pub­lish­ing},
  title = {MINDS: A Trans­la­tor to Embed Math­e­mat­i­cal Expres­sions Inside SPARQL Queries},
  year = 2020
  }
  

  %0 Con­fer­ence Paper
  %1 graux2020-minds
  %A Graux, Damien
  %A Sejdiu, Gez­im
  %A Stadler, Claus
  %A Napoli­tano, Giulio
  %A Lehmann, Jens
  %B Seman­tic Sys­tems. In the Era of Knowl­edge Graphs
  %C Cham
  %D 2020
  %E Blomqvist, Eva
  %E Groth, Paul
  %E de Boer, Vic­tor
  %E Pel­le­gri­ni, Tas­si­lo
  %E Alam, Mehwish
  %E K{ä}fer, Tobias
  %E Kiese­berg, Peter
  %E Kir­rane, Sab­ri­na
  %E Mero{\~{n}}o‑Pe{\~{n}}uela, Albert
  %E Pan­dit, Harsh­vard­han J.
  %I Springer Inter­na­tion­al Pub­lish­ing
  %P 104–117
  %T MINDS: A Trans­la­tor to Embed Math­e­mat­i­cal Expres­sions Inside SPARQL Queries
  %U https://svn.aksw.org/papers/2020/semantics_minds/public.pdf
  %X The recent deploy­ments of seman­tic web tools and the expan­sion of avail­able linked datasets have giv­en users the oppor­tu­ni­ty of build­ing increas­ing­ly com­plex appli­ca­tions. These emerg­ing use cas­es often require queries con­tain­ing math­e­mat­i­cal for­mu­las such as euclid­ean dis­tances or unit con­ver­sions. Cur­rent­ly, the lat­est SPARQL stan­dard (ver­sion 1.1) only embeds basic math oper­a­tors. Thus, to address this short­com­ing, some pop­u­lar SPARQL eval­u­a­tors pro­vide built-in tools to cov­er spe­cif­ic needs; how­ev­er, such tools are not stan­dard yet. To offer users a more gener­ic solu­tion, we pro­pose and share MINDS, a trans­la­tor of math­e­mat­i­cal expres­sions into SPAR­QL-com­pli­ant bind­ings which can be under­stood by any eval­u­a­tor. MINDS there­by facil­i­tates the query design when­ev­er math­e­mat­i­cal com­pu­ta­tions are need­ed in a SPARQL query.
  %@ 978–3‑030–59833‑4
  

• Auto­mat­ic Sub­ject Index­ing with Knowl­edge GraphsIn: LASCAR Work­shop at the Extend­ed Seman­tic Web Con­fer­ence (ESWC)Lisa Wenige, Claus Stadler, Simon Bin, Lorenz Büh­mann, Kurt Jung­hanns and Michael Martin

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{wenige2020kindex,
  author = {Wenige, Lisa and Stadler, Claus and Bin, Simon and Büh­mann, Lorenz and Jung­hanns, Kurt and Mar­tin, Michael},
  book­ti­tle = {LASCAR Work­shop at the Extend­ed Seman­tic Web Con­fer­ence (ESWC)},
  key­words = {sys:relevantFor:infai},
  title = {Auto­mat­ic Sub­ject Index­ing with Knowl­edge Graphs},
  year = 2020
  }
  

  %0 Con­fer­ence Paper
  %1 wenige2020kindex
  %A Wenige, Lisa
  %A Stadler, Claus
  %A Bin, Simon
  %A Büh­mann, Lorenz
  %A Jung­hanns, Kurt
  %A Mar­tin, Michael
  %B LASCAR Work­shop at the Extend­ed Seman­tic Web Con­fer­ence (ESWC)
  %D 2020
  %T Auto­mat­ic Sub­ject Index­ing with Knowl­edge Graphs
  %U https://svn.aksw.org/papers/2020/LASCAR_Kindex/public.pdf
  

• Schema-agnos­tic SPAR­QL-dri­ven faceted search bench­mark gen­er­a­tionIn: Jour­nal of Web Seman­tics, p. 100614Claus Stadler, Simon Bin, Lisa Wenige, Lorenz Büh­mann and Jens Lehmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  In this work, we present a schema-agnos­tic faceted brows­ing bench­mark gen­er­a­tion frame­work for RDF data and SPARQL engines. Faceted search is a tech­nique that allows nar­row­ing down sets of infor­ma­tion items by apply­ing con­straints over their prop­er­ties, where­as facets cor­re­spond to prop­er­ties of these items. While our work can be used to realise real-world faceted search user inter­faces, our focus lies on the con­struc­tion and bench­mark­ing of faceted search queries over knowl­edge graphs. The RDF mod­el exhibits sev­er­al traits that seem­ing­ly make it a nat­ur­al foun­da­tion for faceted search: all infor­ma­tion items are rep­re­sent­ed as RDF resources, prop­er­ty val­ues typ­i­cal­ly already cor­re­spond to mean­ing­ful seman­tic clas­si­fi­ca­tions, and with SPARQL there is a stan­dard lan­guage for uni­form­ly query­ing instance and schema infor­ma­tion. How­ev­er, although faceted search is ubiq­ui­tous today, it is typ­i­cal­ly not per­formed on the RDF mod­el direct­ly. Two major sources of con­cern are the com­plex­i­ty of query gen­er­a­tion and the query per­for­mance. To over­come the for­mer, our frame­work comes with an inter­me­di­ate domain-spe­cif­ic lan­guage. There­by our approach is SPAR­QL-dri­ven which means that every faceted search infor­ma­tion need is inten­sion­al­ly expressed as a sin­gle SPARQL query. In regard to the lat­ter, we inves­ti­gate the pos­si­bil­i­ties and lim­its of real-time SPAR­QL-dri­ven faceted search on con­tem­po­rary triple stores. We report on our find­ings by eval­u­at­ing sys­tems per­for­mance and cor­rect­ness char­ac­ter­is­tics when exe­cut­ing a bench­mark gen­er­at­ed using our gen­er­a­tion frame­work. All com­po­nents, name­ly the bench­mark gen­er­a­tor, the bench­mark run­ners and the under­ly­ing faceted search frame­work, are pub­lished freely avail­able as open source.

  @article{stadler2020facete,
  abstract = {In this work, we present a schema-agnos­tic faceted brows­ing bench­mark gen­er­a­tion frame­work for RDF data and SPARQL engines. Faceted search is a tech­nique that allows nar­row­ing down sets of infor­ma­tion items by apply­ing con­straints over their prop­er­ties, where­as facets cor­re­spond to prop­er­ties of these items. While our work can be used to realise real-world faceted search user inter­faces, our focus lies on the con­struc­tion and bench­mark­ing of faceted search queries over knowl­edge graphs. The RDF mod­el exhibits sev­er­al traits that seem­ing­ly make it a nat­ur­al foun­da­tion for faceted search: all infor­ma­tion items are rep­re­sent­ed as RDF resources, prop­er­ty val­ues typ­i­cal­ly already cor­re­spond to mean­ing­ful seman­tic clas­si­fi­ca­tions, and with SPARQL there is a stan­dard lan­guage for uni­form­ly query­ing instance and schema infor­ma­tion. How­ev­er, although faceted search is ubiq­ui­tous today, it is typ­i­cal­ly not per­formed on the RDF mod­el direct­ly. Two major sources of con­cern are the com­plex­i­ty of query gen­er­a­tion and the query per­for­mance. To over­come the for­mer, our frame­work comes with an inter­me­di­ate domain-spe­cif­ic lan­guage. There­by our approach is SPAR­QL-dri­ven which means that every faceted search infor­ma­tion need is inten­sion­al­ly expressed as a sin­gle SPARQL query. In regard to the lat­ter, we inves­ti­gate the pos­si­bil­i­ties and lim­its of real-time SPAR­QL-dri­ven faceted search on con­tem­po­rary triple stores. We report on our find­ings by eval­u­at­ing sys­tems per­for­mance and cor­rect­ness char­ac­ter­is­tics when exe­cut­ing a bench­mark gen­er­at­ed using our gen­er­a­tion frame­work. All com­po­nents, name­ly the bench­mark gen­er­a­tor, the bench­mark run­ners and the under­ly­ing faceted search frame­work, are pub­lished freely avail­able as open source.},
  author = {Stadler, Claus and Bin, Simon and Wenige, Lisa and Büh­mann, Lorenz and Lehmann, Jens},
  jour­nal = {Jour­nal of Web Seman­tics},
  key­words = {sys:relevantFor:infai},
  pages = 100614,
  title = {Schema-agnos­tic SPAR­QL-dri­ven faceted search bench­mark gen­er­a­tion},
  year = 2020
  }
  

  %0 Jour­nal Arti­cle
  %1 stadler2020facete
  %A Stadler, Claus
  %A Bin, Simon
  %A Wenige, Lisa
  %A Büh­mann, Lorenz
  %A Lehmann, Jens
  %D 2020
  %J Jour­nal of Web Seman­tics
  %P 100614
  %R https://doi.org/10.1016/j.websem.2020.100614
  %T Schema-agnos­tic SPAR­QL-dri­ven faceted search bench­mark gen­er­a­tion
  %U https://svn.aksw.org/papers/2020/JWS_Faceted_Search_Benchmark/public.pdf
  %X In this work, we present a schema-agnos­tic faceted brows­ing bench­mark gen­er­a­tion frame­work for RDF data and SPARQL engines. Faceted search is a tech­nique that allows nar­row­ing down sets of infor­ma­tion items by apply­ing con­straints over their prop­er­ties, where­as facets cor­re­spond to prop­er­ties of these items. While our work can be used to realise real-world faceted search user inter­faces, our focus lies on the con­struc­tion and bench­mark­ing of faceted search queries over knowl­edge graphs. The RDF mod­el exhibits sev­er­al traits that seem­ing­ly make it a nat­ur­al foun­da­tion for faceted search: all infor­ma­tion items are rep­re­sent­ed as RDF resources, prop­er­ty val­ues typ­i­cal­ly already cor­re­spond to mean­ing­ful seman­tic clas­si­fi­ca­tions, and with SPARQL there is a stan­dard lan­guage for uni­form­ly query­ing instance and schema infor­ma­tion. How­ev­er, although faceted search is ubiq­ui­tous today, it is typ­i­cal­ly not per­formed on the RDF mod­el direct­ly. Two major sources of con­cern are the com­plex­i­ty of query gen­er­a­tion and the query per­for­mance. To over­come the for­mer, our frame­work comes with an inter­me­di­ate domain-spe­cif­ic lan­guage. There­by our approach is SPAR­QL-dri­ven which means that every faceted search infor­ma­tion need is inten­sion­al­ly expressed as a sin­gle SPARQL query. In regard to the lat­ter, we inves­ti­gate the pos­si­bil­i­ties and lim­its of real-time SPAR­QL-dri­ven faceted search on con­tem­po­rary triple stores. We report on our find­ings by eval­u­at­ing sys­tems per­for­mance and cor­rect­ness char­ac­ter­is­tics when exe­cut­ing a bench­mark gen­er­at­ed using our gen­er­a­tion frame­work. All com­po­nents, name­ly the bench­mark gen­er­a­tor, the bench­mark run­ners and the under­ly­ing faceted search frame­work, are pub­lished freely avail­able as open source.
  

2019

• RDF-based Deploy­ment Pipelin­ing for Effi­cient Dataset Release Man­age­mentIn: Pro­ceed­ings of the Posters and Demos Track of the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems co-locat­ed with the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMANTICS’19). Karl­sruhe, Ger­manyClaus Stadler, Lisa Wenige, Sebas­t­ian Tramp, Kurt Jung­hanns and Michael Martin

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Open Data por­tals often strug­gle to pro­vide release fea­tures (i.e., sta­ble ver­sion­ing, up-to-date down­load links, rich meta­da­ta des­critions) for their datasets. By this means, wide adop­tion of pub­licly avail­able datasets is hin­dered, since con­sum­ing appli­ca­tions can­not access fresh data sources or might break due to data qual­i­ty issues. While there exists a vari­ety of tools to effi­cient­ly con­trol release process­es in soft­ware devel­op­ment, the man­age­ment of dataset releas­es is not as clear. This paper pro­pos­es a deploy­ment pipeline for effi­cient dataset releas­es that is based on auto­mat­ed enrich­ment of DCAT/DATAID meta­da­ta and is a first step towards effi­cient deploy­ment pipelin­ing for Open Data publishing.

  @inproceedings{stadler-n-2019–rdfdeploy,
  abstract = {Open Data por­tals often strug­gle to pro­vide release fea­tures (i.e., sta­ble ver­sion­ing, up-to-date down­load links, rich meta­da­ta des­critions) for their datasets. By this means, wide adop­tion of pub­licly avail­able datasets is hin­dered, since con­sum­ing appli­ca­tions can­not access fresh data sources or might break due to data qual­i­ty issues. While there exists a vari­ety of tools to effi­cient­ly con­trol release process­es in soft­ware devel­op­ment, the man­age­ment of dataset releas­es is not as clear. This paper pro­pos­es a deploy­ment pipeline for effi­cient dataset releas­es that is based on auto­mat­ed enrich­ment of DCAT/DATAID meta­da­ta and is a first step towards effi­cient deploy­ment pipelin­ing for Open Data pub­lish­ing.},
  address = {Karl­sruhe, Ger­many},
  author = {Stadler, Claus and Wenige, Lisa and Tramp, Sebas­t­ian and Jung­hanns, Kurt and Mar­tin, Michael},
  book­ti­tle = {Pro­ceed­ings of the Posters and Demos Track of the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems co-locat­ed with the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMANTICS’19)},
  key­words = {sys:relevantFor:infai},
  title = {RDF-based Deploy­ment Pipelin­ing for Effi­cient Dataset Release Man­age­ment},
  year = 2019
  }
  

  %0 Con­fer­ence Paper
  %1 stadler-n-2019–rdfdeploy
  %A Stadler, Claus
  %A Wenige, Lisa
  %A Tramp, Sebas­t­ian
  %A Jung­hanns, Kurt
  %A Mar­tin, Michael
  %B Pro­ceed­ings of the Posters and Demos Track of the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems co-locat­ed with the 14th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMANTICS’19)
  %C Karl­sruhe, Ger­many
  %D 2019
  %T RDF-based Deploy­ment Pipelin­ing for Effi­cient Dataset Release Man­age­ment
  %U https://svn.aksw.org/papers/2019/semantics_rdf_pipeline/public.pdf
  %X Open Data por­tals often strug­gle to pro­vide release fea­tures (i.e., sta­ble ver­sion­ing, up-to-date down­load links, rich meta­da­ta des­critions) for their datasets. By this means, wide adop­tion of pub­licly avail­able datasets is hin­dered, since con­sum­ing appli­ca­tions can­not access fresh data sources or might break due to data qual­i­ty issues. While there exists a vari­ety of tools to effi­cient­ly con­trol release process­es in soft­ware devel­op­ment, the man­age­ment of dataset releas­es is not as clear. This paper pro­pos­es a deploy­ment pipeline for effi­cient dataset releas­es that is based on auto­mat­ed enrich­ment of DCAT/DATAID meta­da­ta and is a first step towards effi­cient deploy­ment pipelin­ing for Open Data publishing.
  

2018

• {SQCFrame­work}: Gen­er­at­ing SPARQL Query Con­tain­ment Bench­marks using the SQCFrame­workIn: Demo at ISWCMuham­mad Saleem, Claus Stadler, Qais­er Mehmood, Jens Lehmann and Axel-Cyrille Ngonga Ngomo

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{Saleem2018SQCDmo,
  author = {Saleem, Muham­mad and Stadler, Claus and Mehmood, Qais­er and Lehmann, Jens and Ngo­mo, Axel-Cyrille Ngonga},
  book­ti­tle = {Demo at ISWC},
  key­words = {sys:relevantFor:infai},
  title = {{SQCFrame­work}: Gen­er­at­ing SPARQL Query Con­tain­ment Bench­marks using the SQCFrame­work},
  year = 2018
  }
  

  %0 Con­fer­ence Paper
  %1 Saleem2018SQCDmo
  %A Saleem, Muham­mad
  %A Stadler, Claus
  %A Mehmood, Qais­er
  %A Lehmann, Jens
  %A Ngo­mo, Axel-Cyrille Ngonga
  %B Demo at ISWC
  %D 2018
  %T {SQCFrame­work}: Gen­er­at­ing SPARQL Query Con­tain­ment Bench­marks using the SQCFrame­work
  %U https://svn.aksw.org/papers/2018/ISWC-Demo-SQCFramework/public.pdf
  

2017

• The Tale of Sansa SparkIn: Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence, Poster \& DemosIvan Ermilov, Jens Lehmann, Gez­im Sejdiu, Lorenz Büh­mann, Patrick West­phal, Claus Stadler, Simon Bin, Nilesh Chakraborty, Hen­ning Pet­z­ka, et al.

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{iermilov-2017-sansa-iswc-demo,
  author = {Ermilov, Ivan and Lehmann, Jens and Sejdiu, Gez­im and Büh­mann, Lorenz and West­phal, Patrick and Stadler, Claus and Bin, Simon and Chakraborty, Nilesh and Pet­z­ka, Hen­ning and Saleem, Muham­mad and Ngonga, Axel-Cyrille Ngo­mo and Jabeen, Haji­ra},
  book­ti­tle = {Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence, Poster \& Demos},
  key­words = {buehmann},
  title = {The Tale of Sansa Spark},
  year = 2017
  }
  

  %0 Con­fer­ence Paper
  %1 ier­milov-2017-sansa-iswc-demo
  %A Ermilov, Ivan
  %A Lehmann, Jens
  %A Sejdiu, Gez­im
  %A Büh­mann, Lorenz
  %A West­phal, Patrick
  %A Stadler, Claus
  %A Bin, Simon
  %A Chakraborty, Nilesh
  %A Pet­z­ka, Hen­ning
  %A Saleem, Muham­mad
  %A Ngonga, Axel-Cyrille Ngo­mo
  %A Jabeen, Haji­ra
  %B Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence, Poster \& Demos
  %D 2017
  %T The Tale of Sansa Spark
  %U https://svn.aksw.org/papers/2017/ISWC_SANSA_Demo/public.pdf
  

• Dis­trib­uted Seman­tic Ana­lyt­ics using the SANSA StackIn: Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence — Resources Track (ISWC’2017) : Springer, pp. 147–155Jens Lehmann, Gez­im Sejdiu, Lorenz Büh­mann, Patrick West­phal, Claus Stadler, Ivan Ermilov, Simon Bin, Nilesh Chakraborty, Muham­mad Saleem, et al.

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{lehmann-2017-sansa-iswc,
  author = {Lehmann, Jens and Sejdiu, Gez­im and Büh­mann, Lorenz and West­phal, Patrick and Stadler, Claus and Ermilov, Ivan and Bin, Simon and Chakraborty, Nilesh and Saleem, Muham­mad and Ngonga, Axel-Cyrille Ngo­mo and Jabeen, Haji­ra},
  book­ti­tle = {Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence — Resources Track (ISWC’2017)},
  key­words = {buehmann},
  pages = {147–155},
  pub­lish­er = {Springer},
  title = {Dis­trib­uted Seman­tic Ana­lyt­ics using the SANSA Stack},
  year = 2017
  }
  

  %0 Con­fer­ence Paper
  %1 lehmann-2017-sansa-iswc
  %A Lehmann, Jens
  %A Sejdiu, Gez­im
  %A Büh­mann, Lorenz
  %A West­phal, Patrick
  %A Stadler, Claus
  %A Ermilov, Ivan
  %A Bin, Simon
  %A Chakraborty, Nilesh
  %A Saleem, Muham­mad
  %A Ngonga, Axel-Cyrille Ngo­mo
  %A Jabeen, Haji­ra
  %B Pro­ceed­ings of 16th Inter­na­tion­al Seman­tic Web Con­fer­ence — Resources Track (ISWC’2017)
  %D 2017
  %I Springer
  %P 147–155
  %T Dis­trib­uted Seman­tic Ana­lyt­ics using the SANSA Stack
  %U http://svn.aksw.org/papers/2017/ISWC_SANSA_SoftwareFramework/public.pdf
  

• Bench­mark­ing Faceted Brows­ing Capa­bil­i­ties of Triple­storesIn: 13th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMAN­TiCS 2017), Sep­tem­ber 11–14 2017, Ams­ter­dam, Nether­landsHen­ning Pet­z­ka, Claus Stadler, Geor­gios Kat­sim­pras, Bas­t­ian Haar­mann and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{petzka-semantics-facets,
  author = {Pet­z­ka, Hen­ning and Stadler, Claus and Kat­sim­pras, Geor­gios and Haar­mann, Bas­t­ian and Lehmann, Jens},
  book­ti­tle = {13th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMAN­TiCS 2017), Sep­tem­ber 11–14 2017, Ams­ter­dam, Nether­lands},
  key­words = {sys:relevantFor:infai},
  title = {Bench­mark­ing Faceted Brows­ing Capa­bil­i­ties of Triple­stores},
  year = 2017
  }
  

  %0 Con­fer­ence Paper
  %1 pet­z­ka-seman­tics-facets
  %A Pet­z­ka, Hen­ning
  %A Stadler, Claus
  %A Kat­sim­pras, Geor­gios
  %A Haar­mann, Bas­t­ian
  %A Lehmann, Jens
  %B 13th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems (SEMAN­TiCS 2017), Sep­tem­ber 11–14 2017, Ams­ter­dam, Nether­lands
  %D 2017
  %T Bench­mark­ing Faceted Brows­ing Capa­bil­i­ties of Triple­stores
  %U https://svn.aksw.org/papers/2017/Semantics_faceted_browsing/public.pdf
  

• {SQCFrame­work}: {SPARQL} Query Con­tain­ment Bench­mark Gen­er­a­tion Frame­workIn: K‑CAP 2017: Knowl­edge Cap­ture Con­fer­ence : ACMMuham­mad Saleem, Claus Stadler, Qais­er Mehmood, Jens Lehmann and Axel-Cyrille Ngonga Ngomo

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{Saleem2017SQC,
  author = {Saleem, Muham­mad and Stadler, Claus and Mehmood, Qais­er and Lehmann, Jens and Ngo­mo, Axel-Cyrille Ngonga},
  book­ti­tle = {K‑CAP 2017: Knowl­edge Cap­ture Con­fer­ence},
  key­words = {sys:relevantFor:infai},
  orga­ni­za­tion = {ACM},
  title = {{SQCFrame­work}: {SPARQL} Query Con­tain­ment Bench­mark Gen­er­a­tion Frame­work},
  year = 2017
  }
  

  %0 Con­fer­ence Paper
  %1 Saleem2017SQC
  %A Saleem, Muham­mad
  %A Stadler, Claus
  %A Mehmood, Qais­er
  %A Lehmann, Jens
  %A Ngo­mo, Axel-Cyrille Ngonga
  %B K‑CAP 2017: Knowl­edge Cap­ture Con­fer­ence
  %D 2017
  %T {SQCFrame­work}: {SPARQL} Query Con­tain­ment Bench­mark Gen­er­a­tion Frame­work
  %U https://svn.aksw.org/papers/2017/KCAP_SQCFrameWork/public.pdf
  

• JPA Cri­te­ria Queries over RDF DataIn: Work­shop on Query­ing the Web of Data co-locat­ed with the Extend­ed Seman­tic Web Con­fer­enceClaus Stadler and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{quweda_jpa,
  author = {Stadler, Claus and Lehmann, Jens},
  book­ti­tle = {Work­shop on Query­ing the Web of Data co-locat­ed with the Extend­ed Seman­tic Web Con­fer­ence},
  key­words = {MOLE},
  title = {JPA Cri­te­ria Queries over RDF Data},
  year = 2017
  }
  

  %0 Con­fer­ence Paper
  %1 quweda_jpa
  %A Stadler, Claus
  %A Lehmann, Jens
  %B Work­shop on Query­ing the Web of Data co-locat­ed with the Extend­ed Seman­tic Web Con­fer­ence
  %D 2017
  %T JPA Cri­te­ria Queries over RDF Data
  %U https://svn.aksw.org/papers/2017/QuWeDa_jpa/public.pdf
  

2016

• Towards Sus­tain­able view-based Extract-Trans­form-Load (ETL) Fusion of Open DataIn: Pro­ceed­ings of the Work­shop on 3rd Work­shop on Linked Data Qual­i­ty co-locat­ed with the Euro­pean Seman­tic Web Con­fer­ence 2016 (ESWC 2016)kmueller stadler singh hellmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Open­ly avail­able datasets orig­i­nate from dif­fer­ent data providers which range from gov­ern­ment agen­cies, over com­mer­cial enter­pris­es to com­mu­ni­ties of data enthu­si­asts. Inte­grat­ing dif­fer­ent source datasets into a sin­gle RDF graph by using ETL (Extract-Trans­form-Load) sys- tems which per­form offline trans­for­ma­tion, ontol­ogy match­ing and link- ing tech­niques usu­al­ly takes many iter­a­tions of revi­sions until the tar­get dataset is made free of the most obvi­ous map­ping, link­ing and con­sis- ten­cy errors. Since ETL sys­tems pro­duce the RDF offline, any map- ping or con­tent change requires a re-ingest of the rel­e­vant source data. When deal­ing with het­ero­ge­neous source datasets, cre­at­ing a uni­fied tar- get dataset can be a tedious under­tak­ing. There­fore the paper pro­pos­es an RDF view based inges­tion approach, which allows real-time “debug- ging” of the uni­fied dataset where map­pings and links can be changed with imme­di­ate effect. Once the uni­fied graph pass­es all data qual­i­ty tests, the RDF can be mate­ri­al­ized. This process pos­es an alter­na­tive to exist­ing ETL solutions.

  @inproceedings{kmueller-ldq16-etl-views,
  abstract = {Open­ly avail­able datasets orig­i­nate from dif­fer­ent data providers which range from gov­ern­ment agen­cies, over com­mer­cial enter­pris­es to com­mu­ni­ties of data enthu­si­asts. Inte­grat­ing dif­fer­ent source datasets into a sin­gle RDF graph by using ETL (Extract-Trans­form-Load) sys- tems which per­form offline trans­for­ma­tion, ontol­ogy match­ing and link- ing tech­niques usu­al­ly takes many iter­a­tions of revi­sions until the tar­get dataset is made free of the most obvi­ous map­ping, link­ing and con­sis- ten­cy errors. Since ETL sys­tems pro­duce the RDF offline, any map- ping or con­tent change requires a re-ingest of the rel­e­vant source data. When deal­ing with het­ero­ge­neous source datasets, cre­at­ing a uni­fied tar- get dataset can be a tedious under­tak­ing. There­fore the paper pro­pos­es an RDF view based inges­tion approach, which allows real-time “debug- ging” of the uni­fied dataset where map­pings and links can be changed with imme­di­ate effect. Once the uni­fied graph pass­es all data qual­i­ty tests, the RDF can be mate­ri­al­ized. This process pos­es an alter­na­tive to exist­ing ETL solu­tions.},
  author = {kmueller stadler singh hell­mann},
  book­ti­tle = {Pro­ceed­ings of the Work­shop on 3rd Work­shop on Linked Data Qual­i­ty co-locat­ed with the Euro­pean Seman­tic Web Con­fer­ence 2016 (ESWC 2016)},
  key­words = {hell­mann},
  month = {05},
  title = {Towards Sus­tain­able view-based Extract-Trans­form-Load (ETL) Fusion of Open Data},
  year = 2016
  }
  

  %0 Con­fer­ence Paper
  %1 kmueller-ldq16-etl-views
  %A hell­mann, kmueller stadler singh
  %B Pro­ceed­ings of the Work­shop on 3rd Work­shop on Linked Data Qual­i­ty co-locat­ed with the Euro­pean Seman­tic Web Con­fer­ence 2016 (ESWC 2016)
  %D 2016
  %T Towards Sus­tain­able view-based Extract-Trans­form-Load (ETL) Fusion of Open Data
  %U http://svn.aksw.org/papers/2016/ESWC_ETL_VIEWS/public.pdf
  %X Open­ly avail­able datasets orig­i­nate from dif­fer­ent data providers which range from gov­ern­ment agen­cies, over com­mer­cial enter­pris­es to com­mu­ni­ties of data enthu­si­asts. Inte­grat­ing dif­fer­ent source datasets into a sin­gle RDF graph by using ETL (Extract-Trans­form-Load) sys- tems which per­form offline trans­for­ma­tion, ontol­ogy match­ing and link- ing tech­niques usu­al­ly takes many iter­a­tions of revi­sions until the tar­get dataset is made free of the most obvi­ous map­ping, link­ing and con­sis- ten­cy errors. Since ETL sys­tems pro­duce the RDF offline, any map- ping or con­tent change requires a re-ingest of the rel­e­vant source data. When deal­ing with het­ero­ge­neous source datasets, cre­at­ing a uni­fied tar- get dataset can be a tedious under­tak­ing. There­fore the paper pro­pos­es an RDF view based inges­tion approach, which allows real-time “debug- ging” of the uni­fied dataset where map­pings and links can be changed with imme­di­ate effect. Once the uni­fied graph pass­es all data qual­i­ty tests, the RDF can be mate­ri­al­ized. This process pos­es an alter­na­tive to exist­ing ETL solutions.
  

2015

• Sim­pli­fied {RDB2RDF} Map­pingIn: Pro­ceed­ings of the 8th Work­shop on Linked Data on the Web (LDOW2015), Flo­rence, ItalyClaus Stadler, Joerg Unbe­hauen, Patrick West­phal, Mohamed Ahmed Sherif and Jens Lehmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  The com­bi­na­tion of the advan­tages of wide­ly used rela­tion­al data­bas­es and seman­tic tech­nolo­gies has attract­ed sig­nif­i­cant research over the past decade. In par­tic­u­lar, map­ping lan­guages for the con­ver­sion of data­bas­es to RDF knowl­edge bases have been devel­oped and stan­dard­ized in the form of R2RML. In this arti­cle, we first review those map­ping lan­guages and then devise work towards a uni­fied for­mal mod­el for them. Based on this, we present the Spar­qli­fi­ca­tion Map­ping Lan­guage (SML), which pro­vides an intu­itive way to declare map­pings based on SQL VIEWS and SPARQL con­struct queries. We show that SML has the same expres­siv­i­ty as R2RML by enu­mer­at­ing the lan­guage fea­tures and show the cor­re­spon­dences, and we out­line how one syn­tax can be con­vert­ed into the oth­er. A con­duct­ed user study for this paper jux­ta­pos­ing SML and R2RML pro­vides evi­dence that SML is a more com­pact syn­tax which is eas­i­er to under­stand and read and thus low­ers the bar­ri­er to offer SPARQL access to rela­tion­al databases.

  @inproceedings{sml,
  abstract = {The com­bi­na­tion of the advan­tages of wide­ly used rela­tion­al data­bas­es and seman­tic tech­nolo­gies has attract­ed sig­nif­i­cant research over the past decade. In par­tic­u­lar, map­ping lan­guages for the con­ver­sion of data­bas­es to RDF knowl­edge bases have been devel­oped and stan­dard­ized in the form of R2RML. In this arti­cle, we first review those map­ping lan­guages and then devise work towards a uni­fied for­mal mod­el for them. Based on this, we present the Spar­qli­fi­ca­tion Map­ping Lan­guage (SML), which pro­vides an intu­itive way to declare map­pings based on SQL VIEWS and SPARQL con­struct queries. We show that SML has the same expres­siv­i­ty as R2RML by enu­mer­at­ing the lan­guage fea­tures and show the cor­re­spon­dences, and we out­line how one syn­tax can be con­vert­ed into the oth­er. A con­duct­ed user study for this paper jux­ta­pos­ing SML and R2RML pro­vides evi­dence that SML is a more com­pact syn­tax which is eas­i­er to under­stand and read and thus low­ers the bar­ri­er to offer SPARQL access to rela­tion­al data­bas­es.},
  author = {Stadler, Claus and Unbe­hauen, Joerg and West­phal, Patrick and Sherif, Mohamed Ahmed and Lehmann, Jens},
  book­ti­tle = {Pro­ceed­ings of the 8th Work­shop on Linked Data on the Web (LDOW2015), Flo­rence, Italy},
  key­words = {MOLE},
  title = {Sim­pli­fied {RDB2RDF} Map­ping},
  year = 2015
  }
  

  %0 Con­fer­ence Paper
  %1 sml
  %A Stadler, Claus
  %A Unbe­hauen, Joerg
  %A West­phal, Patrick
  %A Sherif, Mohamed Ahmed
  %A Lehmann, Jens
  %B Pro­ceed­ings of the 8th Work­shop on Linked Data on the Web (LDOW2015), Flo­rence, Italy
  %D 2015
  %T Sim­pli­fied {RDB2RDF} Map­ping
  %U http://svn.aksw.org/papers/2015/LDOW_SML/paper-camery-ready_public.pdf
  %X The com­bi­na­tion of the advan­tages of wide­ly used rela­tion­al data­bas­es and seman­tic tech­nolo­gies has attract­ed sig­nif­i­cant research over the past decade. In par­tic­u­lar, map­ping lan­guages for the con­ver­sion of data­bas­es to RDF knowl­edge bases have been devel­oped and stan­dard­ized in the form of R2RML. In this arti­cle, we first review those map­ping lan­guages and then devise work towards a uni­fied for­mal mod­el for them. Based on this, we present the Spar­qli­fi­ca­tion Map­ping Lan­guage (SML), which pro­vides an intu­itive way to declare map­pings based on SQL VIEWS and SPARQL con­struct queries. We show that SML has the same expres­siv­i­ty as R2RML by enu­mer­at­ing the lan­guage fea­tures and show the cor­re­spon­dences, and we out­line how one syn­tax can be con­vert­ed into the oth­er. A con­duct­ed user study for this paper jux­ta­pos­ing SML and R2RML pro­vides evi­dence that SML is a more com­pact syn­tax which is eas­i­er to under­stand and read and thus low­ers the bar­ri­er to offer SPARQL access to rela­tion­al databases.
  

• The Geo­Know Gen­er­a­tor Work­bench: An Inte­gra­tion Plat­form for Geospa­tial DataIn: Pro­ceed­ings of the 3rd Inter­na­tion­al Work­shop on Seman­tic Web Enter­prise Adop­tion and Best Prac­ticeAle­jan­dra Gar­cia-Rojas, Daniel Hlad­ky, Matthias Wauer, Robert Ise­le, Claus Stadler and Jens Lehmann

  Bib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{wasabi_generator,
  author = {Gar­cia-Rojas, Ale­jan­dra and Hlad­ky, Daniel and Wauer, Matthias and Ise­le, Robert and Stadler, Claus and Lehmann, Jens},
  book­ti­tle = {Pro­ceed­ings of the 3rd Inter­na­tion­al Work­shop on Seman­tic Web Enter­prise Adop­tion and Best Prac­tice},
  key­words = {ontol­ogy},
  title = {The Geo­Know Gen­er­a­tor Work­bench: An Inte­gra­tion Plat­form for Geospa­tial Data},
  year = 2015
  }
  

  %0 Con­fer­ence Paper
  %1 wasabi_generator
  %A Gar­cia-Rojas, Ale­jan­dra
  %A Hlad­ky, Daniel
  %A Wauer, Matthias
  %A Ise­le, Robert
  %A Stadler, Claus
  %A Lehmann, Jens
  %B Pro­ceed­ings of the 3rd Inter­na­tion­al Work­shop on Seman­tic Web Enter­prise Adop­tion and Best Prac­tice
  %D 2015
  %T The Geo­Know Gen­er­a­tor Work­bench: An Inte­gra­tion Plat­form for Geospa­tial Data
  

• Cube­Viz — Explo­ration and Visu­al­iza­tion of Sta­tis­ti­cal Linked DataIn: Pro­ceed­ings of the 24th Inter­na­tion­al Con­fer­ence on World Wide Web, WWW 2015Michael Mar­tin, Kon­rad Abicht, Claus Stadler, S{ö}ren Auer, Axel‑C. Ngonga Ngo­mo and Tom­ma­so Soru

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Cube­Viz is a flex­i­ble explo­ration and visu­al­iza­tion plat­form for sta­tis­ti­cal data rep­re­sent­ed adher­ing to the RDF Data Cube vocab­u­lary. If sta­tis­ti­cal data is pro­vid­ed adher­ing to the Data Cube vocab­u­lary, Cube­Viz exhibits a faceted brows­ing wid­get allow­ing to inter­ac­tive­ly fil­ter obser­va­tions to be visu­al­ized in charts. Based on the select­ed struc­tur­al part, Cube­Viz offers suit­able chart types and options for con­fig­ur­ing the visu­al­iza­tion by users. In this demo we present the Cube­Viz visu­al­iza­tion archi­tec­ture and com­po­nents, sketch its under­ly­ing API and the libraries used to gen­er­ate the desired out­put. By employ­ing advanced intro­spec­tion, analy­sis and visu­al­iza­tion boot­strap­ping tech­niques Cube­Viz hides the schema com­plex­i­ty of the encod­ed data in order to sup­port a user-friend­ly explo­ration experience.

  @inproceedings{martin-www-2015-demo-cubeviz,
  abstract = {Cube­Viz is a flex­i­ble explo­ration and visu­al­iza­tion plat­form for sta­tis­ti­cal data rep­re­sent­ed adher­ing to the RDF Data Cube vocab­u­lary. If sta­tis­ti­cal data is pro­vid­ed adher­ing to the Data Cube vocab­u­lary, Cube­Viz exhibits a faceted brows­ing wid­get allow­ing to inter­ac­tive­ly fil­ter obser­va­tions to be visu­al­ized in charts. Based on the select­ed struc­tur­al part, Cube­Viz offers suit­able chart types and options for con­fig­ur­ing the visu­al­iza­tion by users. In this demo we present the Cube­Viz visu­al­iza­tion archi­tec­ture and com­po­nents, sketch its under­ly­ing API and the libraries used to gen­er­ate the desired out­put. By employ­ing advanced intro­spec­tion, analy­sis and visu­al­iza­tion boot­strap­ping tech­niques Cube­Viz hides the schema com­plex­i­ty of the encod­ed data in order to sup­port a user-friend­ly explo­ration expe­ri­ence.},
  author = {Mar­tin, Michael and Abicht, Kon­rad and Stadler, Claus and Auer, S{ö}ren and Ngonga Ngo­mo, Axel‑C. and Soru, Tom­ma­so},
  book­ti­tle = {Pro­ceed­ings of the 24th Inter­na­tion­al Con­fer­ence on World Wide Web, WWW 2015},
  key­words = {topic_Exploration},
  title = {Cube­Viz — Explo­ration and Visu­al­iza­tion of Sta­tis­ti­cal Linked Data},
  year = 2015
  }
  

  %0 Con­fer­ence Paper
  %1 mar­tin-www-2015-demo-cube­viz
  %A Mar­tin, Michael
  %A Abicht, Kon­rad
  %A Stadler, Claus
  %A Auer, S{ö}ren
  %A Ngonga Ngo­mo, Axel‑C.
  %A Soru, Tom­ma­so
  %B Pro­ceed­ings of the 24th Inter­na­tion­al Con­fer­ence on World Wide Web, WWW 2015
  %D 2015
  %T Cube­Viz — Explo­ration and Visu­al­iza­tion of Sta­tis­ti­cal Linked Data
  %U https://svn.aksw.org/papers/2015/WWW_Demo_CubeViz/public.pdf
  %X Cube­Viz is a flex­i­ble explo­ration and visu­al­iza­tion plat­form for sta­tis­ti­cal data rep­re­sent­ed adher­ing to the RDF Data Cube vocab­u­lary. If sta­tis­ti­cal data is pro­vid­ed adher­ing to the Data Cube vocab­u­lary, Cube­Viz exhibits a faceted brows­ing wid­get allow­ing to inter­ac­tive­ly fil­ter obser­va­tions to be visu­al­ized in charts. Based on the select­ed struc­tur­al part, Cube­Viz offers suit­able chart types and options for con­fig­ur­ing the visu­al­iza­tion by users. In this demo we present the Cube­Viz visu­al­iza­tion archi­tec­ture and com­po­nents, sketch its under­ly­ing API and the libraries used to gen­er­ate the desired out­put. By employ­ing advanced intro­spec­tion, analy­sis and visu­al­iza­tion boot­strap­ping tech­niques Cube­Viz hides the schema com­plex­i­ty of the encod­ed data in order to sup­port a user-friend­ly explo­ration experience.
  

• RDF Edit­ing on the WebIn: SEMANTICS 2015, SEM ’15. Vien­na, Aus­tria : ACMClaus Stadler, Natanael Arndt, Michael Mar­tin and Jens Lehmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  While sev­er­al tools for sim­pli­fy­ing the task of visu­al­iz­ing (SPARQL acces­si­ble) RDF data on the Web are avail­able today, there is a lack of cor­re­spond­ing tools for exploit­ing stan­dard HTML forms direct­ly for RDF edit­ing. The few relat­ed exist­ing sys­tems rough­ly fall in the cat­e­gories of (a) appli­ca­tions that are not aimed at being reused as com­po­nents, (b) form gen­er­a­tors, which auto­mat­i­cal­ly cre­ate forms from a giv­en schema — pos­si­bly derived from instance data — or © form tem­plate proces­sors which cre­ate forms from a man­u­al­ly cre­at­ed spec­i­fi­ca­tion. Fur­ther­more, these sys­tems usu­al­ly come with their own wid­get library, which can only be extend­ed by wrap­ping exist­ing wid­gets. In this paper, we present the Angu­lar­JS-based \emph{Rdf Edit eXten­sion} (REX) sys­tem, which facil­i­tates the enhance­ment of stan­dard HTML forms as well as many exist­ing Angu­lar­JS wid­gets with RDF edit­ing sup­port by means of a set of HTML attrib­ut­es. We demon­strate our sys­tem though the real­iza­tion of sev­er­al usage scenarios.

  @inproceedings{rex_pd,
  abstract = {While sev­er­al tools for sim­pli­fy­ing the task of visu­al­iz­ing (SPARQL acces­si­ble) RDF data on the Web are avail­able today, there is a lack of cor­re­spond­ing tools for exploit­ing stan­dard HTML forms direct­ly for RDF edit­ing. The few relat­ed exist­ing sys­tems rough­ly fall in the cat­e­gories of (a) appli­ca­tions that are not aimed at being reused as com­po­nents, (b) form gen­er­a­tors, which auto­mat­i­cal­ly cre­ate forms from a giv­en schema — pos­si­bly derived from instance data — or © form tem­plate proces­sors which cre­ate forms from a man­u­al­ly cre­at­ed spec­i­fi­ca­tion. Fur­ther­more, these sys­tems usu­al­ly come with their own wid­get library, which can only be extend­ed by wrap­ping exist­ing wid­gets. In this paper, we present the Angu­lar­JS-based \emph{Rdf Edit eXten­sion} (REX) sys­tem, which facil­i­tates the enhance­ment of stan­dard HTML forms as well as many exist­ing Angu­lar­JS wid­gets with RDF edit­ing sup­port by means of a set of HTML attrib­ut­es. We demon­strate our sys­tem though the real­iza­tion of sev­er­al usage sce­nar­ios.},
  author = {Stadler, Claus and Arndt, Natanael and Mar­tin, Michael and Lehmann, Jens},
  book­ti­tle = {SEMANTICS 2015},
  key­words = {sys:relevantFor:infai},
  month = {09},
  pub­lish­er = {ACM},
  series = {SEM ’15},
  title = {RDF Edit­ing on the Web},
  year = 2015
  }
  

  %0 Con­fer­ence Paper
  %1 rex_pd
  %A Stadler, Claus
  %A Arndt, Natanael
  %A Mar­tin, Michael
  %A Lehmann, Jens
  %B SEMANTICS 2015
  %D 2015
  %I ACM
  %T RDF Edit­ing on the Web
  %U http://ceur-ws.org/Vol-1481/paper29.pdf
  %X While sev­er­al tools for sim­pli­fy­ing the task of visu­al­iz­ing (SPARQL acces­si­ble) RDF data on the Web are avail­able today, there is a lack of cor­re­spond­ing tools for exploit­ing stan­dard HTML forms direct­ly for RDF edit­ing. The few relat­ed exist­ing sys­tems rough­ly fall in the cat­e­gories of (a) appli­ca­tions that are not aimed at being reused as com­po­nents, (b) form gen­er­a­tors, which auto­mat­i­cal­ly cre­ate forms from a giv­en schema — pos­si­bly derived from instance data — or © form tem­plate proces­sors which cre­ate forms from a man­u­al­ly cre­at­ed spec­i­fi­ca­tion. Fur­ther­more, these sys­tems usu­al­ly come with their own wid­get library, which can only be extend­ed by wrap­ping exist­ing wid­gets. In this paper, we present the Angu­lar­JS-based \emph{Rdf Edit eXten­sion} (REX) sys­tem, which facil­i­tates the enhance­ment of stan­dard HTML forms as well as many exist­ing Angu­lar­JS wid­gets with RDF edit­ing sup­port by means of a set of HTML attrib­ut­es. We demon­strate our sys­tem though the real­iza­tion of sev­er­al usage scenarios.
  

2014

• Jas­sa — {A} JavaScript suite for SPAR­QL-based faceted searchIn: Pro­ceed­ings of the {ISWC} Devel­op­ers Work­shop 2014, co-locat­ed with the 13th Inter­na­tion­al Seman­tic Web Con­fer­ence {(ISWC} 2014), Riva del Gar­da, Italy, Octo­ber 19, 2014., pp. 31–36Claus Stadler, Patrick West­phal and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{jassa,
  author = {Stadler, Claus and West­phal, Patrick and Lehmann, Jens},
  book­ti­tle = {Pro­ceed­ings of the {ISWC} Devel­op­ers Work­shop 2014, co-locat­ed with the 13th Inter­na­tion­al Seman­tic Web Con­fer­ence {(ISWC} 2014), Riva del Gar­da, Italy, Octo­ber 19, 2014.},
  cross­ref = {DBLP:conf/semweb/2014dev},
  key­words = {sys:relevantFor:infai},
  pages = {31–36},
  title = {Jas­sa — {A} JavaScript suite for SPAR­QL-based faceted search},
  year = 2014
  }
  

  %0 Con­fer­ence Paper
  %1 jas­sa
  %A Stadler, Claus
  %A West­phal, Patrick
  %A Lehmann, Jens
  %B Pro­ceed­ings of the {ISWC} Devel­op­ers Work­shop 2014, co-locat­ed with the 13th Inter­na­tion­al Seman­tic Web Con­fer­ence {(ISWC} 2014), Riva del Gar­da, Italy, Octo­ber 19, 2014.
  %D 2014
  %P 31–36
  %T Jas­sa — {A} JavaScript suite for SPAR­QL-based faceted search
  %U http://ceur-ws.org/Vol-1268/paper6.pdf
  

• LD view­er-linked data pre­sen­ta­tion frame­workIn: Pro­ceed­ings of the 10th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems : ACM, pp. 124–131Denis Lukovnikov, Claus Stadler and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{ld_viewer,
  author = {Lukovnikov, Denis and Stadler, Claus and Lehmann, Jens},
  book­ti­tle = {Pro­ceed­ings of the 10th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems},
  key­words = {sys:relevantFor:infai},
  orga­ni­za­tion = {ACM},
  pages = {124–131},
  title = {LD view­er-linked data pre­sen­ta­tion frame­work},
  year = 2014
  }
  

  %0 Con­fer­ence Paper
  %1 ld_viewer
  %A Lukovnikov, Denis
  %A Stadler, Claus
  %A Lehmann, Jens
  %B Pro­ceed­ings of the 10th Inter­na­tion­al Con­fer­ence on Seman­tic Sys­tems
  %D 2014
  %P 124–131
  %T LD view­er-linked data pre­sen­ta­tion frame­work
  %U https://svn.aksw.org/papers/2014/Semantics_ld_viewer/public.pdf
  

• {Explor­ing the Web of Spa­tial Data with Facete}In: Com­pan­ion pro­ceed­ings of 23rd Inter­na­tion­al World Wide Web Con­fer­ence (WWW), pp. 175–178Claus Stadler, Michael Mar­tin and S{ö}ren Auer

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{stadler-www,
  author = {Stadler, Claus and Mar­tin, Michael and Auer, S{ö}ren},
  book­ti­tle = {Com­pan­ion pro­ceed­ings of 23rd Inter­na­tion­al World Wide Web Con­fer­ence (WWW)},
  key­words = {facete},
  pages = {175–178},
  title = {{Explor­ing the Web of Spa­tial Data with Facete}},
  year = 2014
  }
  

  %0 Con­fer­ence Paper
  %1 stadler-www
  %A Stadler, Claus
  %A Mar­tin, Michael
  %A Auer, S{ö}ren
  %B Com­pan­ion pro­ceed­ings of 23rd Inter­na­tion­al World Wide Web Con­fer­ence (WWW)
  %D 2014
  %P 175–178
  %T {Explor­ing the Web of Spa­tial Data with Facete}
  %U https://svn.aksw.org/papers/2014/WWW_Facete/public.pdf
  

• DBpe­dia View­er — An Inte­gra­tive Inter­face for DBpe­dia lever­ag­ing the DBpe­dia Ser­vice Eco Sys­temIn: Proc. of the Linked Data on the Web 2014 Work­shopDenis Lukovnikov, Claus Stadler, Dim­itris Kon­tokostas, Sebas­t­ian Hell­mann and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{ldow_dbpedia_viewer,
  author = {Lukovnikov, Denis and Stadler, Claus and Kon­tokostas, Dim­itris and Hell­mann, Sebas­t­ian and Lehmann, Jens},
  book­ti­tle = {Proc. of the Linked Data on the Web 2014 Work­shop},
  key­words = {hell­mann},
  title = {DBpe­dia View­er — An Inte­gra­tive Inter­face for DBpe­dia lever­ag­ing the DBpe­dia Ser­vice Eco Sys­tem},
  year = 2014
  }
  

  %0 Con­fer­ence Paper
  %1 ldow_dbpedia_viewer
  %A Lukovnikov, Denis
  %A Stadler, Claus
  %A Kon­tokostas, Dim­itris
  %A Hell­mann, Sebas­t­ian
  %A Lehmann, Jens
  %B Proc. of the Linked Data on the Web 2014 Work­shop
  %D 2014
  %T DBpe­dia View­er — An Inte­gra­tive Inter­face for DBpe­dia lever­ag­ing the DBpe­dia Ser­vice Eco Sys­tem
  %U https://svn.aksw.org/papers/2014/LDOW_DBpediaInterface/public.pdf
  

• The Geo­Know Gen­er­a­tor: Man­ag­ing Geospa­tial Data in the Linked Data WebIn: Pro­ceed­ings of the Link­ing Geospa­tial Data Work­shopJon Jay Le Grange, Jens Lehmann, Spiros Athana­siou, Ale­jan­dra Gar­cia Rojas, Gior­gos Giannopou­los, Daniel Hlad­ky, Robert Ise­le, Axel-Cyrille {Ngonga Ngo­mo}, Mohamed Ahmed Sherif, et al.

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{lgd_geoknow_generator,
  author = {Grange, Jon Jay Le and Lehmann, Jens and Athana­siou, Spiros and Rojas, Ale­jan­dra Gar­cia and Giannopou­los, Gior­gos and Hlad­ky, Daniel and Ise­le, Robert and {Ngonga Ngo­mo}, Axel-Cyrille and Sherif, Mohamed Ahmed and Stadler, Claus and Wauer, Matthias},
  book­ti­tle = {Pro­ceed­ings of the Link­ing Geospa­tial Data Work­shop},
  key­words = {MOLE},
  title = {The Geo­Know Gen­er­a­tor: Man­ag­ing Geospa­tial Data in the Linked Data Web},
  year = 2014
  }
  

  %0 Con­fer­ence Paper
  %1 lgd_geoknow_generator
  %A Grange, Jon Jay Le
  %A Lehmann, Jens
  %A Athana­siou, Spiros
  %A Rojas, Ale­jan­dra Gar­cia
  %A Giannopou­los, Gior­gos
  %A Hlad­ky, Daniel
  %A Ise­le, Robert
  %A {Ngonga Ngo­mo}, Axel-Cyrille
  %A Sherif, Mohamed Ahmed
  %A Stadler, Claus
  %A Wauer, Matthias
  %B Pro­ceed­ings of the Link­ing Geospa­tial Data Work­shop
  %D 2014
  %T The Geo­Know Gen­er­a­tor: Man­ag­ing Geospa­tial Data in the Linked Data Web
  %U http://jens-lehmann.org/files/2014/lgd_geoknow_generator.pdf
  

• Qual­i­ty Assur­ance of RDB2RDF Map­pingsPatrick West­phal, Claus Stadler and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @techreport{rdb2rdf_qa,
  author = {West­phal, Patrick and Stadler, Claus and Lehmann, Jens},
  key­words = {sys:relevantFor:infai},
  title = {Qual­i­ty Assur­ance of RDB2RDF Map­pings},
  year = 2014
  }
  

  %0 Report
  %1 rdb2rdf_qa
  %A West­phal, Patrick
  %A Stadler, Claus
  %A Lehmann, Jens
  %D 2014
  %T Qual­i­ty Assur­ance of RDB2RDF Map­pings
  %U http://svn.aksw.org/papers/2014/report_QA_RDB2RDF/public.pdf
  

• Facil­i­tat­ing the Explo­ration and Visu­al­iza­tion of Linked DataIn: S{ö}ren Auer, Vol­ha Bryl and Sebas­t­ian Tramp (eds.): Linked Open Data—Creating Knowl­edge Out of Inter­linked Data, Lec­ture Notes in Com­put­er Sci­ence : Springer Inter­na­tion­al Pub­lish­ing — ISBN 978–3‑319–09845‑6, pp. 90–107Chris­t­ian Mad­er, Michael Mar­tin and Claus Stadler

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  The cre­ation and the improve­ment of tools that cov­er explorato­ry and visu­al­iza­tion tasks for Linked Data were one of the major goals focused in the LOD2 project. Tools that sup­port those tasks are regard­ed as essen­tial for the Web of Data, since they can act as a user-ori­ent­ed start­ing point for data cus­tomers. Dur­ing the project, sev­er­al devel­op­ment efforts were made, whose results either facil­i­tate the explo­ration and visu­al­iza­tion direct­ly (such as OntoWi­ki, the Piv­ot Brows­er) or can be used to sup­port such tasks. In this chap­ter we present the three select­ed solu­tions rsine, Cube­Viz and Facete.

  @incollection{LOD2Book-ExplorationVisualization,
  abstract = {The cre­ation and the improve­ment of tools that cov­er explorato­ry and visu­al­iza­tion tasks for Linked Data were one of the major goals focused in the LOD2 project. Tools that sup­port those tasks are regard­ed as essen­tial for the Web of Data, since they can act as a user-ori­ent­ed start­ing point for data cus­tomers. Dur­ing the project, sev­er­al devel­op­ment efforts were made, whose results either facil­i­tate the explo­ration and visu­al­iza­tion direct­ly (such as OntoWi­ki, the Piv­ot Brows­er) or can be used to sup­port such tasks. In this chap­ter we present the three select­ed solu­tions rsine, Cube­Viz and Facete.},
  author = {Mad­er, Chris­t­ian and Mar­tin, Michael and Stadler, Claus},
  book­ti­tle = {Linked Open Data—Creating Knowl­edge Out of Inter­linked Data},
  edi­tor = {Auer, S{ö}ren and Bryl, Vol­ha and Tramp, Sebas­t­ian},
  key­words = {facete},
  pages = {90–107},
  pub­lish­er = {Springer Inter­na­tion­al Pub­lish­ing},
  series = {Lec­ture Notes in Com­put­er Sci­ence},
  title = {Facil­i­tat­ing the Explo­ration and Visu­al­iza­tion of Linked Data},
  year = 2014
  }
  

  %0 Book Sec­tion
  %1 LOD2­Book-Explo­rationVi­su­al­iza­tion
  %A Mad­er, Chris­t­ian
  %A Mar­tin, Michael
  %A Stadler, Claus
  %B Linked Open Data—Creating Knowl­edge Out of Inter­linked Data
  %D 2014
  %E Auer, S{ö}ren
  %E Bryl, Vol­ha
  %E Tramp, Sebas­t­ian
  %I Springer Inter­na­tion­al Pub­lish­ing
  %P 90–107
  %R 10.1007/978–3‑319–09846-3_5
  %T Facil­i­tat­ing the Explo­ration and Visu­al­iza­tion of Linked Data
  %U https://svn.aksw.org/papers/2014/LOD_rsine/public.pdf
  %X The cre­ation and the improve­ment of tools that cov­er explorato­ry and visu­al­iza­tion tasks for Linked Data were one of the major goals focused in the LOD2 project. Tools that sup­port those tasks are regard­ed as essen­tial for the Web of Data, since they can act as a user-ori­ent­ed start­ing point for data cus­tomers. Dur­ing the project, sev­er­al devel­op­ment efforts were made, whose results either facil­i­tate the explo­ration and visu­al­iza­tion direct­ly (such as OntoWi­ki, the Piv­ot Brows­er) or can be used to sup­port such tasks. In this chap­ter we present the three select­ed solu­tions rsine, Cube­Viz and Facete.
  %@ 978–3‑319–09845‑6
  

2013

• Gen­er­at­ing SPARQL queries using tem­platesIn: WIAS journal,Vol. 11, No. 3, 2013.Saeedeh Shekar­pour, S{ö}ren Auer, Axel-Cyrille {Ngonga Ngo­mo}, Daniel Ger­ber, Sebas­t­ian Hell­mann and Claus Stadler

  Bib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{SHE+journal2013,
  author = {Shekar­pour, Saeedeh and Auer, S{ö}ren and {Ngonga Ngo­mo}, Axel-Cyrille and Ger­ber, Daniel and Hell­mann, Sebas­t­ian and Stadler, Claus},
  book­ti­tle = {WIAS journal,Vol. 11, No. 3, 2013.},
  key­words = {MOLE},
  title = {Gen­er­at­ing SPARQL queries using tem­plates},
  year = 2013
  }
  

  %0 Con­fer­ence Paper
  %1 SHE+journal2013
  %A Shekar­pour, Saeedeh
  %A Auer, S{ö}ren
  %A {Ngonga Ngo­mo}, Axel-Cyrille
  %A Ger­ber, Daniel
  %A Hell­mann, Sebas­t­ian
  %A Stadler, Claus
  %B WIAS journal,Vol. 11, No. 3, 2013.
  %D 2013
  %T Gen­er­at­ing SPARQL queries using templates
  

• Extrak­tion, Map­ping und Ver­linkung von Dat­en im WebIn: Daten­bank Spek­trum vol. 13, Nr. 2, pp. 77–87S{ö}ren Auer, Jens Lehmann, Axel-Cyrille Ngonga Ngo­mo, Claus Stadler and J{ö}rg Unbehauen

  URLBib­TeXEnd­NoteBib­Son­o­my

  @article{AUE+13a,
  author = {Auer, S{ö}ren and Lehmann, Jens and Ngo­mo, Axel-Cyrille Ngonga and Stadler, Claus and Unbe­hauen, J{ö}rg},
  jour­nal = {Daten­bank Spek­trum},
  key­words = {MOLE},
  num­ber = 2,
  pages = {77–87},
  title = {Extrak­tion, Map­ping und Ver­linkung von Dat­en im Web},
  vol­ume = 13,
  year = 2013
  }
  

  %0 Jour­nal Arti­cle
  %1 AUE+13a
  %A Auer, S{ö}ren
  %A Lehmann, Jens
  %A Ngo­mo, Axel-Cyrille Ngonga
  %A Stadler, Claus
  %A Unbe­hauen, J{ö}rg
  %D 2013
  %J Daten­bank Spek­trum
  %N 2
  %P 77–87
  %T Extrak­tion, Map­ping und Ver­linkung von Dat­en im Web
  %U http://jens-lehmann.org/files/2013/db_spektrum_linked_data.pdf
  %V 13
  

• CSV2RDF: User-Dri­ven CSV to RDF Mass Con­ver­sion Frame­workIn: Pro­ceed­ings of the ISEM ’13, Sep­tem­ber 04 — 06 2013, Graz, Aus­triaIvan Ermilov, S{ö}ren Auer and Claus Stadler

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Gov­ern­ments and pub­lic admin­is­tra­tions start­ed recent­ly to pub­lish large amounts of struc­tured data on the Web, most­ly in the form of tab­u­lar data such as CSV files or Excel sheets. Var­i­ous tools and projects have been launched aim­ing at facil­i­tat­ing the lift­ing of tab­u­lar data to reach seman­ti­cal­ly struc­tured and linked data. How­ev­er, none of these tools sup­port­ed a tru­ly incre­men­tal, pay-as-you-go data pub­li­ca­tion and map­ping strat­e­gy, which enables effort shar­ing between data own­ers, com­mu­ni­ty experts and con­sumers. In this arti­cle, we present an approach for enabling the user-dri­ven seman­tic map­ping of large amounts tab­u­lar data. We devise a sim­ple map­ping lan­guage for tab­u­lar data, which is easy to under­stand even for casu­al users, but expres­sive enough to cov­er the vast major­i­ty of poten­tial tab­u­lar map­pings use cas­es. We out­line a for­mal approach for map­ping tab­u­lar data to RDF. Default map­pings are auto­mat­i­cal­ly cre­at­ed and can be revised by the com­mu­ni­ty using a seman­tic wiki. The map­pings are exe­cut­ed using a sophis­ti­cat­ed stream­ing RDB2RDF con­ver­sion. We report about the deploy­ment of our approach at the Pan-Euro­pean data por­tal PublicData.eu, where we trans­formed and enriched almost 10,000 datasets account­ing for 7.3 bil­lion triples.

  @inproceedings{ermilov-ivan-2013-isem,
  abstract = {Gov­ern­ments and pub­lic admin­is­tra­tions start­ed recent­ly to pub­lish large amounts of struc­tured data on the Web, most­ly in the form of tab­u­lar data such as CSV files or Excel sheets. Var­i­ous tools and projects have been launched aim­ing at facil­i­tat­ing the lift­ing of tab­u­lar data to reach seman­ti­cal­ly struc­tured and linked data. How­ev­er, none of these tools sup­port­ed a tru­ly incre­men­tal, pay-as-you-go data pub­li­ca­tion and map­ping strat­e­gy, which enables effort shar­ing between data own­ers, com­mu­ni­ty experts and con­sumers. In this arti­cle, we present an approach for enabling the user-dri­ven seman­tic map­ping of large amounts tab­u­lar data. We devise a sim­ple map­ping lan­guage for tab­u­lar data, which is easy to under­stand even for casu­al users, but expres­sive enough to cov­er the vast major­i­ty of poten­tial tab­u­lar map­pings use cas­es. We out­line a for­mal approach for map­ping tab­u­lar data to RDF. Default map­pings are auto­mat­i­cal­ly cre­at­ed and can be revised by the com­mu­ni­ty using a seman­tic wiki. The map­pings are exe­cut­ed using a sophis­ti­cat­ed stream­ing RDB2RDF con­ver­sion. We report about the deploy­ment of our approach at the Pan-Euro­pean data por­tal PublicData.eu, where we trans­formed and enriched almost 10,000 datasets account­ing for 7.3 bil­lion triples.},
  author = {Ermilov, Ivan and Auer, S{ö}ren and Stadler, Claus},
  book­ti­tle = {Pro­ceed­ings of the ISEM ’13, Sep­tem­ber 04 — 06 2013, Graz, Aus­tria},
  key­words = {sys:relevantFor:infai},
  title = {CSV2RDF: User-Dri­ven CSV to RDF Mass Con­ver­sion Frame­work},
  year = 2013
  }
  

  %0 Con­fer­ence Paper
  %1 ermilov-ivan-2013-isem
  %A Ermilov, Ivan
  %A Auer, S{ö}ren
  %A Stadler, Claus
  %B Pro­ceed­ings of the ISEM ’13, Sep­tem­ber 04 — 06 2013, Graz, Aus­tria
  %D 2013
  %T CSV2RDF: User-Dri­ven CSV to RDF Mass Con­ver­sion Frame­work
  %U http://svn.aksw.org/papers/2013/ISemantics_CSV2RDF/public.pdf
  %X Gov­ern­ments and pub­lic admin­is­tra­tions start­ed recent­ly to pub­lish large amounts of struc­tured data on the Web, most­ly in the form of tab­u­lar data such as CSV files or Excel sheets. Var­i­ous tools and projects have been launched aim­ing at facil­i­tat­ing the lift­ing of tab­u­lar data to reach seman­ti­cal­ly struc­tured and linked data. How­ev­er, none of these tools sup­port­ed a tru­ly incre­men­tal, pay-as-you-go data pub­li­ca­tion and map­ping strat­e­gy, which enables effort shar­ing between data own­ers, com­mu­ni­ty experts and con­sumers. In this arti­cle, we present an approach for enabling the user-dri­ven seman­tic map­ping of large amounts tab­u­lar data. We devise a sim­ple map­ping lan­guage for tab­u­lar data, which is easy to under­stand even for casu­al users, but expres­sive enough to cov­er the vast major­i­ty of poten­tial tab­u­lar map­pings use cas­es. We out­line a for­mal approach for map­ping tab­u­lar data to RDF. Default map­pings are auto­mat­i­cal­ly cre­at­ed and can be revised by the com­mu­ni­ty using a seman­tic wiki. The map­pings are exe­cut­ed using a sophis­ti­cat­ed stream­ing RDB2RDF con­ver­sion. We report about the deploy­ment of our approach at the Pan-Euro­pean data por­tal PublicData.eu, where we trans­formed and enriched almost 10,000 datasets account­ing for 7.3 bil­lion triples.
  

• Increas­ing the Finan­cial Trans­paren­cy of Euro­pean Com­mis­sion Project Fund­ingIn: Seman­tic Web Jour­nal vol. Spe­cial Call for Linked Dataset descrip­tions, Nr. 2, pp. 157–164Michael Mar­tin, Claus Stadler, Philipp Frischmuth and Jens Lehmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  The Finan­cial Trans­paren­cy Sys­tem (FTS) of the Euro­pean Com­mis­sion con­tains infor­ma­tion about grants for Euro­pean Union projects start­ing from 2007. It allows users to get an overview on EU fund­ing, includ­ing infor­ma­tion on ben­e­fi­cia­ries as well as the amount and type of expen­di­ture and infor­ma­tion on the respon­si­ble EU depart­ment. The orig­i­nal dataset is freely avail­able on the Euro­pean Com­mis­sion web­site, where users can query the data using an HTML form and down­load it in CSV and most recent­ly XML for­mat. In this arti­cle, we describe the trans­for­ma­tion of this data to RDF and its inter­link­ing with oth­er datasets. We show that this allows inter­est­ing queries over the data, which were very dif­fi­cult with­out this con­ver­sion. The main ben­e­fit of the dataset is an increased finan­cial trans­paren­cy of EU project fund­ing. The RDF ver­sion of the FTS dataset will become part of the EU Open Data Por­tal and even­tu­al­ly be host­ed and main­tained by the Euro­pean Union itself.

  @article{martin-fts,
  abstract = {The Finan­cial Trans­paren­cy Sys­tem (FTS) of the Euro­pean Com­mis­sion con­tains infor­ma­tion about grants for Euro­pean Union projects start­ing from 2007. It allows users to get an overview on EU fund­ing, includ­ing infor­ma­tion on ben­e­fi­cia­ries as well as the amount and type of expen­di­ture and infor­ma­tion on the respon­si­ble EU depart­ment. The orig­i­nal dataset is freely avail­able on the Euro­pean Com­mis­sion web­site, where users can query the data using an HTML form and down­load it in CSV and most recent­ly XML for­mat. In this arti­cle, we describe the trans­for­ma­tion of this data to RDF and its inter­link­ing with oth­er datasets. We show that this allows inter­est­ing queries over the data, which were very dif­fi­cult with­out this con­ver­sion. The main ben­e­fit of the dataset is an increased finan­cial trans­paren­cy of EU project fund­ing. The RDF ver­sion of the FTS dataset will become part of the EU Open Data Por­tal and even­tu­al­ly be host­ed and main­tained by the Euro­pean Union itself.},
  author = {Mar­tin, Michael and Stadler, Claus and Frischmuth, Philipp and Lehmann, Jens},
  jour­nal = {Seman­tic Web Jour­nal},
  key­words = {MOLE},
  num­ber = 2,
  pages = {157–164},
  title = {Increas­ing the Finan­cial Trans­paren­cy of Euro­pean Com­mis­sion Project Fund­ing},
  vol­ume = {Spe­cial Call for Linked Dataset descrip­tions},
  year = 2013
  }
  

  %0 Jour­nal Arti­cle
  %1 mar­tin-fts
  %A Mar­tin, Michael
  %A Stadler, Claus
  %A Frischmuth, Philipp
  %A Lehmann, Jens
  %D 2013
  %J Seman­tic Web Jour­nal
  %N 2
  %P 157–164
  %T Increas­ing the Finan­cial Trans­paren­cy of Euro­pean Com­mis­sion Project Fund­ing
  %U http://www.semantic-web-journal.net/system/files/swj435.pdf
  %V Spe­cial Call for Linked Dataset descrip­tions
  %X The Finan­cial Trans­paren­cy Sys­tem (FTS) of the Euro­pean Com­mis­sion con­tains infor­ma­tion about grants for Euro­pean Union projects start­ing from 2007. It allows users to get an overview on EU fund­ing, includ­ing infor­ma­tion on ben­e­fi­cia­ries as well as the amount and type of expen­di­ture and infor­ma­tion on the respon­si­ble EU depart­ment. The orig­i­nal dataset is freely avail­able on the Euro­pean Com­mis­sion web­site, where users can query the data using an HTML form and down­load it in CSV and most recent­ly XML for­mat. In this arti­cle, we describe the trans­for­ma­tion of this data to RDF and its inter­link­ing with oth­er datasets. We show that this allows inter­est­ing queries over the data, which were very dif­fi­cult with­out this con­ver­sion. The main ben­e­fit of the dataset is an increased finan­cial trans­paren­cy of EU project fund­ing. The RDF ver­sion of the FTS dataset will become part of the EU Open Data Por­tal and even­tu­al­ly be host­ed and main­tained by the Euro­pean Union itself.
  

2012

• Nav­i­ga­tion-induced Knowl­edge Engi­neer­ing by Exam­pleIn: JISTSebas­t­ian Hell­mann, Jens Lehmann, J{ö}rg Unbe­hauen, Claus Stadler, Thanh Nghia Lam and Markus Strohmaier

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{hellmann-jist-2012-NKE,
  author = {Hell­mann, Sebas­t­ian and Lehmann, Jens and Unbe­hauen, J{ö}rg and Stadler, Claus and Lam, Thanh Nghia and Strohmaier, Markus},
  book­ti­tle = {JIST},
  key­words = {MOLE},
  title = {Nav­i­ga­tion-induced Knowl­edge Engi­neer­ing by Exam­ple},
  year = 2012
  }
  

  %0 Con­fer­ence Paper
  %1 hell­mann-jist-2012-NKE
  %A Hell­mann, Sebas­t­ian
  %A Lehmann, Jens
  %A Unbe­hauen, J{ö}rg
  %A Stadler, Claus
  %A Lam, Thanh Nghia
  %A Strohmaier, Markus
  %B JIST
  %D 2012
  %T Nav­i­ga­tion-induced Knowl­edge Engi­neer­ing by Exam­ple
  %U http://svn.aksw.org/papers/2012/JIST_NKE/public.pdf
  

• Access­ing Rela­tion­al Data on the Web with Spar­qlMapIn: JISTJ{ö}rg Unbe­hauen, Claus Stadler and S{ö}ren Auer

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{unbehauen-jist-2012-sparqlmap,
  author = {Unbe­hauen, J{ö}rg and Stadler, Claus and Auer, S{ö}ren},
  book­ti­tle = {JIST},
  key­words = {sys:relevantFor:infai},
  title = {Access­ing Rela­tion­al Data on the Web with Spar­qlMap},
  year = 2012
  }
  

  %0 Con­fer­ence Paper
  %1 unbe­hauen-jist-2012-spar­qlmap
  %A Unbe­hauen, J{ö}rg
  %A Stadler, Claus
  %A Auer, S{ö}ren
  %B JIST
  %D 2012
  %T Access­ing Rela­tion­al Data on the Web with Spar­qlMap
  %U https://svn.aksw.org/papers/2012/JIST_SparqlMap/public.pdf
  

• Linked­Ge­o­Da­ta: A Core for a Web of Spa­tial Open DataIn: Seman­tic Web Jour­nal vol. 3, Nr. 4, pp. 333–354Claus Stadler, Jens Lehmann, Kon­rad H{ö}ffner and S{ö}ren Auer

  URLBib­TeXEnd­NoteBib­Son­o­my

  @article{SLHA11,
  author = {Stadler, Claus and Lehmann, Jens and H{ö}ffner, Kon­rad and Auer, S{ö}ren},
  jour­nal = {Seman­tic Web Jour­nal},
  key­words = {sys:relevantFor:infai},
  num­ber = 4,
  pages = {333–354},
  title = {Linked­Ge­o­Da­ta: A Core for a Web of Spa­tial Open Data},
  vol­ume = 3,
  year = 2012
  }
  

  %0 Jour­nal Arti­cle
  %1 SLHA11
  %A Stadler, Claus
  %A Lehmann, Jens
  %A H{ö}ffner, Kon­rad
  %A Auer, S{ö}ren
  %D 2012
  %J Seman­tic Web Jour­nal
  %N 4
  %P 333–354
  %T Linked­Ge­o­Da­ta: A Core for a Web of Spa­tial Open Data
  %U http://jens-lehmann.org/files/2012/linkedgeodata2.pdf
  %V 3
  

• The Ger­man {DBpe­dia}: A Sense Repos­i­to­ry for Link­ing Enti­tiesIn: Sebas­t­ian Hell­mann, Claus Stadler and Jens Lehmann

  Bib­TeXEnd­NoteBib­Son­o­my

  @incollection{Hellmann2012GermanDBpedia,
  author = {Hell­mann, Sebas­t­ian and Stadler, Claus and Lehmann, Jens},
  cross­ref = {Springer-ldl},
  key­words = {hell­mann},
  title = {The Ger­man {DBpe­dia}: A Sense Repos­i­to­ry for Link­ing Enti­ties},
  year = 2012
  }
  

  %0 Book Sec­tion
  %1 Hellmann2012GermanDBpedia
  %A Hell­mann, Sebas­t­ian
  %A Stadler, Claus
  %A Lehmann, Jens
  %D 2012
  %T The Ger­man {DBpe­dia}: A Sense Repos­i­to­ry for Link­ing Entities
  

• {DB}pedia and the {L}ive {E}xtraction of {S}tructured {D}ata from {W}ikipediaIn: Pro­gram: elec­tron­ic library and infor­ma­tion sys­tems vol. 46, p. 27Mohamed Morsey, Jens Lehmann, S{ö}ren Auer, Claus Stadler and Sebas­t­ian Hellmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Pur­pose — DBpe­dia extracts struc­tured infor­ma­tion from Wikipedia, inter­links it with oth­er knowl­edge bases and freely pub­lish­es the results on the Web using Linked Data and SPARQL. How­ev­er, the DBpe­dia release process is heavy-weight and releas­es are some­times based on sev­er­al months old data. DBpe­dia-Live solves this prob­lem by pro­vid­ing a live syn­chro­niza­tion method based on the update stream of Wikipedia. Design/methodology/approach — Wikipedia pro­vides DBpe­dia with a con­tin­u­ous stream of updates, i.e. a stream of recent­ly updat­ed arti­cles. DBpe­dia-Live process­es that stream on the fly to obtain RDF data and stores the extract­ed data back to DBpe­dia. DBpe­dia-Live pub­lish­es the new­ly added/deleted triples in files, in order to enable syn­chro­niza­tion between our DBpe­dia end­point and oth­er DBpe­dia mir­rors. Find­ings — Dur­ing the real­iza­tion of DBpe­dia-Live we learned, that it is cru­cial to process Wikipedia updates in a pri­or­i­ty queue. Recent­ly-updat­ed Wikipedia arti­cles should have the high­est pri­or­i­ty, over map­ping-changes and unmod­i­fied pages. An over­all find­ing is that there is a plen­ty of oppor­tu­ni­ties aris­ing from the emerg­ing Web of Data for librar­i­ans. Prac­ti­cal impli­ca­tions — DBpe­dia had and has a great effect on the Web of Data and became a crys­tal­liza­tion point for it. Many com­pa­nies and researchers use DBpe­dia and its pub­lic ser­vices to improve their appli­ca­tions and research approach­es. The DBpe­dia-Live frame­work improves DBpe­dia fur­ther by time­ly syn­chro­niz­ing it with Wikipedia, which is rel­e­vant for many use cas­es requir­ing up-to-date infor­ma­tion. Originality/value — The new DBpe­dia-Live frame­work adds new fea­tures to the old DBpe­dia-Live frame­work, e.g. abstract extrac­tion, ontol­ogy changes, and change­sets publication.

  @article{dbpedia_live_2012,
  abstract = {Pur­pose — DBpe­dia extracts struc­tured infor­ma­tion from Wikipedia, inter­links it with oth­er knowl­edge bases and freely pub­lish­es the results on the Web using Linked Data and SPARQL. How­ev­er, the DBpe­dia release process is heavy-weight and releas­es are some­times based on sev­er­al months old data. DBpe­dia-Live solves this prob­lem by pro­vid­ing a live syn­chro­niza­tion method based on the update stream of Wikipedia. Design/methodology/approach — Wikipedia pro­vides DBpe­dia with a con­tin­u­ous stream of updates, i.e. a stream of recent­ly updat­ed arti­cles. DBpe­dia-Live process­es that stream on the fly to obtain RDF data and stores the extract­ed data back to DBpe­dia. DBpe­dia-Live pub­lish­es the new­ly added/deleted triples in files, in order to enable syn­chro­niza­tion between our DBpe­dia end­point and oth­er DBpe­dia mir­rors. Find­ings — Dur­ing the real­iza­tion of DBpe­dia-Live we learned, that it is cru­cial to process Wikipedia updates in a pri­or­i­ty queue. Recent­ly-updat­ed Wikipedia arti­cles should have the high­est pri­or­i­ty, over map­ping-changes and unmod­i­fied pages. An over­all find­ing is that there is a plen­ty of oppor­tu­ni­ties aris­ing from the emerg­ing Web of Data for librar­i­ans. Prac­ti­cal impli­ca­tions — DBpe­dia had and has a great effect on the Web of Data and became a crys­tal­liza­tion point for it. Many com­pa­nies and researchers use DBpe­dia and its pub­lic ser­vices to improve their appli­ca­tions and research approach­es. The DBpe­dia-Live frame­work improves DBpe­dia fur­ther by time­ly syn­chro­niz­ing it with Wikipedia, which is rel­e­vant for many use cas­es requir­ing up-to-date infor­ma­tion. Originality/value — The new DBpe­dia-Live frame­work adds new fea­tures to the old DBpe­dia-Live frame­work, e.g. abstract extrac­tion, ontol­ogy changes, and change­sets pub­li­ca­tion.},
  author = {Morsey, Mohamed and Lehmann, Jens and Auer, S{ö}ren and Stadler, Claus and Hell­mann, Sebas­t­ian},
  jour­nal = {Pro­gram: elec­tron­ic library and infor­ma­tion sys­tems},
  key­words = {hell­mann},
  pages = 27,
  title = {{DB}pedia and the {L}ive {E}xtraction of {S}tructured {D}ata from {W}ikipedia},
  vol­ume = 46,
  year = 2012
  }
  

  %0 Jour­nal Arti­cle
  %1 dbpedia_live_2012
  %A Morsey, Mohamed
  %A Lehmann, Jens
  %A Auer, S{ö}ren
  %A Stadler, Claus
  %A Hell­mann, Sebas­t­ian
  %D 2012
  %J Pro­gram: elec­tron­ic library and infor­ma­tion sys­tems
  %P 27
  %T {DB}pedia and the {L}ive {E}xtraction of {S}tructured {D}ata from {W}ikipedia
  %U http://svn.aksw.org/papers/2011/DBpedia_Live/public.pdf
  %V 46
  %X Pur­pose — DBpe­dia extracts struc­tured infor­ma­tion from Wikipedia, inter­links it with oth­er knowl­edge bases and freely pub­lish­es the results on the Web using Linked Data and SPARQL. How­ev­er, the DBpe­dia release process is heavy-weight and releas­es are some­times based on sev­er­al months old data. DBpe­dia-Live solves this prob­lem by pro­vid­ing a live syn­chro­niza­tion method based on the update stream of Wikipedia. Design/methodology/approach — Wikipedia pro­vides DBpe­dia with a con­tin­u­ous stream of updates, i.e. a stream of recent­ly updat­ed arti­cles. DBpe­dia-Live process­es that stream on the fly to obtain RDF data and stores the extract­ed data back to DBpe­dia. DBpe­dia-Live pub­lish­es the new­ly added/deleted triples in files, in order to enable syn­chro­niza­tion between our DBpe­dia end­point and oth­er DBpe­dia mir­rors. Find­ings — Dur­ing the real­iza­tion of DBpe­dia-Live we learned, that it is cru­cial to process Wikipedia updates in a pri­or­i­ty queue. Recent­ly-updat­ed Wikipedia arti­cles should have the high­est pri­or­i­ty, over map­ping-changes and unmod­i­fied pages. An over­all find­ing is that there is a plen­ty of oppor­tu­ni­ties aris­ing from the emerg­ing Web of Data for librar­i­ans. Prac­ti­cal impli­ca­tions — DBpe­dia had and has a great effect on the Web of Data and became a crys­tal­liza­tion point for it. Many com­pa­nies and researchers use DBpe­dia and its pub­lic ser­vices to improve their appli­ca­tions and research approach­es. The DBpe­dia-Live frame­work improves DBpe­dia fur­ther by time­ly syn­chro­niz­ing it with Wikipedia, which is rel­e­vant for many use cas­es requir­ing up-to-date infor­ma­tion. Originality/value — The new DBpe­dia-Live frame­work adds new fea­tures to the old DBpe­dia-Live frame­work, e.g. abstract extrac­tion, ontol­ogy changes, and change­sets publication.
  

• Assess­ing Linked Data Map­pings Using Net­work Mea­suresIn: Pro­ceed­ings of the 9th Extend­ed Seman­tic Web Con­fer­ence, Lec­ture Notes in Com­put­er Sci­ence. vol. 7295 : Springer, pp. 87–102Christophe Gu{é}ret, Paul T. Groth, Claus Stadler and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{Gueret2012,
  author = {Gu{é}ret, Christophe and Groth, Paul T. and Stadler, Claus and Lehmann, Jens},
  book­ti­tle = {Pro­ceed­ings of the 9th Extend­ed Seman­tic Web Con­fer­ence},
  key­words = {sys:relevantFor:infai},
  pages = {87–102},
  pub­lish­er = {Springer},
  series = {Lec­ture Notes in Com­put­er Sci­ence},
  title = {Assess­ing Linked Data Map­pings Using Net­work Mea­sures},
  vol­ume = 7295,
  year = 2012
  }
  

  %0 Con­fer­ence Paper
  %1 Gueret2012
  %A Gu{é}ret, Christophe
  %A Groth, Paul T.
  %A Stadler, Claus
  %A Lehmann, Jens
  %B Pro­ceed­ings of the 9th Extend­ed Seman­tic Web Con­fer­ence
  %D 2012
  %I Springer
  %P 87–102
  %T Assess­ing Linked Data Map­pings Using Net­work Mea­sures
  %U http://jens-lehmann.org/files/2012/linked_mapping_qa.pdf
  %V 7295
  

• {T}owards an {O}pen-{G}overnmental {D}ata {W}ebIn: Pro­ceed­ings of the KESW2012Ivan Ermilov, Claus Stadler, Michael Mar­tin and S{ö}ren Auer

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Up to the present day much effort has been made to pub­lish gov­ern­ment data on the Web. How­ev­er, such data has been pub­lished in dif­fer­ent for­mats. For any par­tic­u­lar source and use (e.g. explo­ration, visu­al­iza­tion, inte­gra­tion) of such infor­ma­tion spe­cial appli­ca­tions have to be writ­ten. This lim­its the over­all usabil­i­ty of the infor­ma­tion pro­vid­ed and makes it dif­fcult to access infor­ma­tion resources. These lim­i­ta­tions can be over­rid­den, if the infor­ma­tion will be pro­vid­ed using a homo­ge­neous data and access mod­el com­ply­ing with the Linked Data prin­ci­ples. In this paper we show­case how raw Open Gov­ern­ment Data (OGD) from het­ero­ge­neous sources can be processed, con­vert­ed, pub­lished and used on the Web of Linked Data. In par­tic­u­lar we demon­strate our expe­ri­ence in pro­cess­ing of OGD on two use cas­es: the Dig­i­tal Agen­da Score­board and the Finan­cial Trans­paren­cy Sys­tem of the Euro­pean Commission.

  @inproceedings{ermilov-i-2012–a,
  abstract = {Up to the present day much effort has been made to pub­lish gov­ern­ment data on the Web. How­ev­er, such data has been pub­lished in dif­fer­ent for­mats. For any par­tic­u­lar source and use (e.g. explo­ration, visu­al­iza­tion, inte­gra­tion) of such infor­ma­tion spe­cial appli­ca­tions have to be writ­ten. This lim­its the over­all usabil­i­ty of the infor­ma­tion pro­vid­ed and makes it dif­fcult to access infor­ma­tion resources. These lim­i­ta­tions can be over­rid­den, if the infor­ma­tion will be pro­vid­ed using a homo­ge­neous data and access mod­el com­ply­ing with the Linked Data prin­ci­ples. In this paper we show­case how raw Open Gov­ern­ment Data (OGD) from het­ero­ge­neous sources can be processed, con­vert­ed, pub­lished and used on the Web of Linked Data. In par­tic­u­lar we demon­strate our expe­ri­ence in pro­cess­ing of OGD on two use cas­es: the Dig­i­tal Agen­da Score­board and the Finan­cial Trans­paren­cy Sys­tem of the Euro­pean Com­mis­sion.},
  author = {Ermilov, Ivan and Stadler, Claus and Mar­tin, Michael and Auer, S{ö}ren},
  book­ti­tle = {Pro­ceed­ings of the KESW2012},
  key­words = {sys:relevantFor:lod2},
  title = {{T}owards an {O}pen-{G}overnmental {D}ata {W}eb},
  year = 2012
  }
  

  %0 Con­fer­ence Paper
  %1 ermilov-i-2012–a
  %A Ermilov, Ivan
  %A Stadler, Claus
  %A Mar­tin, Michael
  %A Auer, S{ö}ren
  %B Pro­ceed­ings of the KESW2012
  %D 2012
  %T {T}owards an {O}pen-{G}overnmental {D}ata {W}eb
  %U http://svn.aksw.org/papers/2012/KESW2012_OpenGovWeb/public.pdf
  %X Up to the present day much effort has been made to pub­lish gov­ern­ment data on the Web. How­ev­er, such data has been pub­lished in dif­fer­ent for­mats. For any par­tic­u­lar source and use (e.g. explo­ration, visu­al­iza­tion, inte­gra­tion) of such infor­ma­tion spe­cial appli­ca­tions have to be writ­ten. This lim­its the over­all usabil­i­ty of the infor­ma­tion pro­vid­ed and makes it dif­fcult to access infor­ma­tion resources. These lim­i­ta­tions can be over­rid­den, if the infor­ma­tion will be pro­vid­ed using a homo­ge­neous data and access mod­el com­ply­ing with the Linked Data prin­ci­ples. In this paper we show­case how raw Open Gov­ern­ment Data (OGD) from het­ero­ge­neous sources can be processed, con­vert­ed, pub­lished and used on the Web of Linked Data. In par­tic­u­lar we demon­strate our expe­ri­ence in pro­cess­ing of OGD on two use cas­es: the Dig­i­tal Agen­da Score­board and the Finan­cial Trans­paren­cy Sys­tem of the Euro­pean Commission.
  

• Man­ag­ing the life-cycle of Linked Data with the {LOD2} StackIn: Pro­ceed­ings of Inter­na­tion­al Seman­tic Web Con­fer­ence (ISWC 2012). — 22\% accep­tance rateS{ö}ren Auer, Lorenz B{ü}hmann, Chris­t­ian Dirschl, Orri Erling, Michael Hausen­blas, Robert Ise­le, Jens Lehmann, Michael Mar­tin, Pablo N. Mendes, et al.

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{Auer+ISWC-2012,
  author = {Auer, S{ö}ren and B{ü}hmann, Lorenz and Dirschl, Chris­t­ian and Erling, Orri and Hausen­blas, Michael and Ise­le, Robert and Lehmann, Jens and Mar­tin, Michael and Mendes, Pablo N. and van Nuffe­len, Bert and Stadler, Claus and Tramp, Sebas­t­ian and Williams, Hugh},
  book­ti­tle = {Pro­ceed­ings of Inter­na­tion­al Seman­tic Web Con­fer­ence (ISWC 2012)},
  key­words = {sys:relevantFor:lod2},
  note = {22\% accep­tance rate},
  title = {Man­ag­ing the life-cycle of Linked Data with the {LOD2} Stack},
  year = 2012
  }
  

  %0 Con­fer­ence Paper
  %1 Auer+ISWC-2012
  %A Auer, S{ö}ren
  %A B{ü}hmann, Lorenz
  %A Dirschl, Chris­t­ian
  %A Erling, Orri
  %A Hausen­blas, Michael
  %A Ise­le, Robert
  %A Lehmann, Jens
  %A Mar­tin, Michael
  %A Mendes, Pablo N.
  %A van Nuffe­len, Bert
  %A Stadler, Claus
  %A Tramp, Sebas­t­ian
  %A Williams, Hugh
  %B Pro­ceed­ings of Inter­na­tion­al Seman­tic Web Con­fer­ence (ISWC 2012)
  %D 2012
  %T Man­ag­ing the life-cycle of Linked Data with the {LOD2} Stack
  %U http://iswc2012.semanticweb.org/sites/default/files/76500001.pdf
  

• Knowl­edge Extrac­tion from Struc­tured SourcesIn: Ste­fano Ceri and Mar­co Bram­bil­la (eds.): Search Com­put­ing — Broad­en­ing Web Search, Lec­ture Notes in Com­put­er Sci­ence. vol. 7538 : Springer, pp. 34–52J{ö}rg Unbe­hauen, Sebas­t­ian Hell­mann, S{ö}ren Auer and Claus Stadler

  URLBib­TeXEnd­NoteBib­Son­o­my

  @incollection{UnbehauenSeCo2012,
  author = {Unbe­hauen, J{ö}rg and Hell­mann, Sebas­t­ian and Auer, S{ö}ren and Stadler, Claus},
  book­ti­tle = {Search Com­put­ing — Broad­en­ing Web Search},
  edi­tor = {Ceri, Ste­fano and Bram­bil­la, Mar­co},
  key­words = {hell­mann},
  pages = {34–52},
  pub­lish­er = {Springer},
  series = {Lec­ture Notes in Com­put­er Sci­ence},
  title = {Knowl­edge Extrac­tion from Struc­tured Sources},
  vol­ume = 7538,
  year = 2012
  }
  

  %0 Book Sec­tion
  %1 UnbehauenSeCo2012
  %A Unbe­hauen, J{ö}rg
  %A Hell­mann, Sebas­t­ian
  %A Auer, S{ö}ren
  %A Stadler, Claus
  %B Search Com­put­ing — Broad­en­ing Web Search
  %D 2012
  %E Ceri, Ste­fano
  %E Bram­bil­la, Mar­co
  %I Springer
  %P 34–52
  %T Knowl­edge Extrac­tion from Struc­tured Sources
  %U http://svn.aksw.org/papers/2012/SearchComputing_KnowledgeExtraction/public.pdf
  %V 7538
  

2011

• Key­word-dri­ven SPARQL Query Gen­er­a­tion Lever­ag­ing Back­ground Knowl­edgeIn: Inter­na­tion­al Con­fer­ence on Web Intel­li­genceSaeedeh Shekar­pour, S{ö}ren Auer, Axel-Cyrille {Ngonga Ngo­mo}, Daniel Ger­ber, Sebas­t­ian Hell­mann and Claus Stadler

  Bib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{SHE+11,
  author = {Shekar­pour, Saeedeh and Auer, S{ö}ren and {Ngonga Ngo­mo}, Axel-Cyrille and Ger­ber, Daniel and Hell­mann, Sebas­t­ian and Stadler, Claus},
  book­ti­tle = {Inter­na­tion­al Con­fer­ence on Web Intel­li­gence},
  key­words = {hell­mann},
  title = {Key­word-dri­ven SPARQL Query Gen­er­a­tion Lever­ag­ing Back­ground Knowl­edge},
  year = 2011
  }
  

  %0 Con­fer­ence Paper
  %1 SHE+11
  %A Shekar­pour, Saeedeh
  %A Auer, S{ö}ren
  %A {Ngonga Ngo­mo}, Axel-Cyrille
  %A Ger­ber, Daniel
  %A Hell­mann, Sebas­t­ian
  %A Stadler, Claus
  %B Inter­na­tion­al Con­fer­ence on Web Intel­li­gence
  %D 2011
  %T Key­word-dri­ven SPARQL Query Gen­er­a­tion Lever­ag­ing Back­ground Knowledge
  

• Linked Data Qual­i­ty Assess­ment through Net­work Analy­sisIn: ISWC 2011 Posters and DemosChristophe Gu{é}ret, Paul Groth, Claus Stadler and Jens Lehmann

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{iswc-11-pd-linkqa,
  author = {Gu{é}ret, Christophe and Groth, Paul and Stadler, Claus and Lehmann, Jens},
  book­ti­tle = {ISWC 2011 Posters and Demos},
  key­words = {sys:relevantFor:infai},
  title = {Linked Data Qual­i­ty Assess­ment through Net­work Analy­sis},
  year = 2011
  }
  

  %0 Con­fer­ence Paper
  %1 iswc-11-pd-linkqa
  %A Gu{é}ret, Christophe
  %A Groth, Paul
  %A Stadler, Claus
  %A Lehmann, Jens
  %B ISWC 2011 Posters and Demos
  %D 2011
  %T Linked Data Qual­i­ty Assess­ment through Net­work Analy­sis
  %U http://jens-lehmann.org/files/2011/iswc_pd_linkqa.pdf
  

2010

• {U}pdate {S}trategies for {DB}pedia {L}iveIn: 6th Work­shop on Script­ing and Devel­op­ment for the Seman­tic Web Colo­cat­ed with ESWC 2010 30th or 31st May, 2010 Crete, GreeceClaus Stadler, Michael Mar­tin, Jens Lehmann and Sebas­t­ian Hellmann

  AbstractURLBib­TeXEnd­NoteBib­Son­o­my

  Wikipedia is one of the largest pub­lic infor­ma­tion spaces with a huge user com­mu­ni­ty, which col­lab­o­ra­tive­ly works on the largest online ency­clo­pe­dia. Their users add or edit up to 150 thou­sand wiki pages per day. The DBpe­dia project extracts RDF from Wikipedia and inter­links it with oth­er knowl­edge bases. In the DBpe­dia live extrac­tion mode, Wikipedia edits are instant­ly processed to update infor­ma­tion in DBpe­dia. Due to the high num­ber of edits and the growth of Wikipedia, the update process has to be very effi­cient and scal­able. In this paper, we present dif­fer­ent strate­gies to tack­le this chal­leng­ing prob­lem and describe how we mod­i­fied the DBpe­dia live extrac­tion algo­rithm to work more efficiently.

  @inproceedings{stadler-c-2010–a,
  abstract = {Wikipedia is one of the largest pub­lic infor­ma­tion spaces with a huge user com­mu­ni­ty, which col­lab­o­ra­tive­ly works on the largest online ency­clo­pe­dia. Their users add or edit up to 150 thou­sand wiki pages per day. The DBpe­dia project extracts RDF from Wikipedia and inter­links it with oth­er knowl­edge bases. In the DBpe­dia live extrac­tion mode, Wikipedia edits are instant­ly processed to update infor­ma­tion in DBpe­dia. Due to the high num­ber of edits and the growth of Wikipedia, the update process has to be very effi­cient and scal­able. In this paper, we present dif­fer­ent strate­gies to tack­le this chal­leng­ing prob­lem and describe how we mod­i­fied the DBpe­dia live extrac­tion algo­rithm to work more effi­cient­ly.},
  author = {Stadler, Claus and Mar­tin, Michael and Lehmann, Jens and Hell­mann, Sebas­t­ian},
  book­ti­tle = {6th Work­shop on Script­ing and Devel­op­ment for the Seman­tic Web Colo­cat­ed with ESWC 2010 30th or 31st May, 2010 Crete, Greece},
  key­words = {MOLE},
  title = {{U}pdate {S}trategies for {DB}pedia {L}ive},
  year = 2010
  }
  

  %0 Con­fer­ence Paper
  %1 stadler-c-2010–a
  %A Stadler, Claus
  %A Mar­tin, Michael
  %A Lehmann, Jens
  %A Hell­mann, Sebas­t­ian
  %B 6th Work­shop on Script­ing and Devel­op­ment for the Seman­tic Web Colo­cat­ed with ESWC 2010 30th or 31st May, 2010 Crete, Greece
  %D 2010
  %T {U}pdate {S}trategies for {DB}pedia {L}ive
  %U http://jens-lehmann.org/files/2010/dbpedia_live_eswc.pdf
  %X Wikipedia is one of the largest pub­lic infor­ma­tion spaces with a huge user com­mu­ni­ty, which col­lab­o­ra­tive­ly works on the largest online ency­clo­pe­dia. Their users add or edit up to 150 thou­sand wiki pages per day. The DBpe­dia project extracts RDF from Wikipedia and inter­links it with oth­er knowl­edge bases. In the DBpe­dia live extrac­tion mode, Wikipedia edits are instant­ly processed to update infor­ma­tion in DBpe­dia. Due to the high num­ber of edits and the growth of Wikipedia, the update process has to be very effi­cient and scal­able. In this paper, we present dif­fer­ent strate­gies to tack­le this chal­leng­ing prob­lem and describe how we mod­i­fied the DBpe­dia live extrac­tion algo­rithm to work more efficiently.
  

2009

• {DBpe­dia} Live Extrac­tionIn: Proc. of 8th Inter­na­tion­al Con­fer­ence on Ontolo­gies, Data­Bas­es, and Appli­ca­tions of Seman­tics (ODBASE), Lec­ture Notes in Com­put­er Sci­ence. vol. 5871, pp. 1209–1223Sebas­t­ian Hell­mann, Claus Stadler, Jens Lehmann and S{ö}ren Auer

  URLBib­TeXEnd­NoteBib­Son­o­my

  @inproceedings{hellmann_odbase_dbpedia_live_09,
  author = {Hell­mann, Sebas­t­ian and Stadler, Claus and Lehmann, Jens and Auer, S{ö}ren},
  book­ti­tle = {Proc. of 8th Inter­na­tion­al Con­fer­ence on Ontolo­gies, Data­Bas­es, and Appli­ca­tions of Seman­tics (ODBASE)},
  key­words = {MOLE},
  pages = {1209–1223},
  series = {Lec­ture Notes in Com­put­er Sci­ence},
  title = {{DBpe­dia} Live Extrac­tion},
  vol­ume = 5871,
  year = 2009
  }
  

  %0 Con­fer­ence Paper
  %1 hellmann_odbase_dbpedia_live_09
  %A Hell­mann, Sebas­t­ian
  %A Stadler, Claus
  %A Lehmann, Jens
  %A Auer, S{ö}ren
  %B Proc. of 8th Inter­na­tion­al Con­fer­ence on Ontolo­gies, Data­Bas­es, and Appli­ca­tions of Seman­tics (ODBASE)
  %D 2009
  %P 1209–1223
  %R doi:10.1007/978–3‑642–05151-7_33
  %T {DBpe­dia} Live Extrac­tion
  %U http://svn.aksw.org/papers/2009/ODBASE_LiveExtraction/dbpedia_live_extraction_public.pdf
  %V 5871