Olivier Pivert is currently a full Professor of Computer Science at the National School of Applied Sciences and Technology, Lannion, France; and a Member of the Institute for Research in Computer Science and Random Systems where he heads the Shaman research team.

• Foreword xiAnne LAURENT and Dominique LAURENTPreface xiiiOlivier PIVERTChapter 1. NoSQL Languages and Systems 1Kim NGUYỄN1.1. Introduction 11.1.1. The rise of NoSQL systems and languages 11.1.2. Overview of NoSQL concepts 41.1.3. Current trends of French research in NoSQL languages 61.2. Join implementations on top of MapReduce 71.3. Models for NoSQL languages and systems 121.4. New challenges for database research 161.5. Bibliography 18Chapter 2. Distributed SPARQL Query Processing: A Case Study with Apache Spark 21Bernd AMANN, Olivier CURÉ and Hubert NAACKE2.1. Introduction 212.2. RDF and SPARQL 222.2.1. RDF framework and data model 222.2.2. SPARQL query language 252.3. SPARQL query processing 292.3.1. SPARQL with and without RDF/S entailment 292.3.2. Query optimization 302.3.3. Triple store systems 332.4. SPARQL and MapReduce 342.4.1. MapReduce-based SPARQL processing 352.4.2. Related work 392.5. SPARQL on Apache Spark 412.5.1. Apache Spark 412.5.2. SPARQL on Spark 422.5.3. Experimental evaluation 482.6. Bibliography 53Chapter 3. Doing Web Data: from Dataset Recommendation to Data Linking 57Manel ACHICHI, Mohamed BEN ELLEFI, Zohra BELLAHSENE and Konstantin TODOROV3.1. Introduction 573.1.1. The Semantic Web vision 573.1.2. Linked data life cycles 583.1.3. Chapter overview 613.2. Datasets recommendation for data linking 623.2.1. Process definition 633.2.2. Dataset recommendation for data linking based on a Semantic Web index 643.2.3. Dataset recommendation for data linking based on social networks 643.2.4. Dataset recommendation for data linking based on domain-specific keywords 653.2.5. Dataset recommendation for data linking based on topic modeling 653.2.6. Dataset recommendation for data linking based on topic profiles 663.2.7. Dataset recommendation for data linking based on intensional profiling 673.2.8. Discussion on dataset recommendation approaches 683.3. Challenges of linking data 693.3.1. Value dimension 703.3.2. Ontological dimension 743.3.3. Logical dimension 773.4. Techniques applied to the data linking process 783.4.1. Data linking techniques 793.4.2. Discussion 833.5. Conclusion 863.6. Bibliography 87Chapter 4. Big Data Integration in Cloud Environments: Requirements, Solutions and Challenges 93Rami SELLAMI and Bruno DEFUDE4.1. Introduction 934.2. Big Data integration requirements in Cloud environments 964.3. Automatic data store selection and discovery 994.3.1. Introduction 994.3.2. Model-based approaches 994.3.3. Matching-oriented approaches 1004.3.4. Comparison 1024.4. Unique access for all data stores 1034.4.1. Introduction 1034.4.2. ODBAPI: A unified REST API for relational and NoSQL data stores 1044.4.3. Other works 1054.4.4. Comparison 1074.5. Unified data model and query languages 1084.5.1. Introduction 1084.5.2. Data models of classical data integration approaches 1094.5.3. A global schema to unify the view over relational and NoSQL data stores 1104.5.4. Other works 1134.5.5. Comparison 1174.6. Query processing and optimization 1184.6.1. Introduction 1184.6.2. Federated query language approaches 1184.6.3. Integrated query language approaches 1214.6.4. Comparison 1244.7. Summary and open issues 1254.7.1. Summary 1254.7.2. Open issues 1274.8. Conclusion 1294.9. Bibliography 129Chapter 5. Querying RDF Data: A Multigraph-based Approach 135Vijay INGALALLI, Dino IENCO and Pascal PONCELET5.1. Introduction 1355.2. Related work 1375.3. Background and preliminaries 1375.3.1. RDF data 1385.3.2. SPARQL query 1405.3.3. SPARQL querying by adopting multigraph homomorphism 1425.4. AMBER: A SPARQL querying engine 1435.5. Index construction 1445.5.1. Attribute index 1445.5.2. Vertex signature index 1455.5.3. Vertex neighborhood index 1485.6. Query matching procedure 1495.6.1. Vertex-level processing 1515.6.2. Processing satellite vertices 1525.6.3. Arbitrary query processing 1545.7. Experimental analysis 1595.7.1. Experimental setup 1595.7.2. Workload generation 1605.7.3. Comparison with RDF engines 1615.8. Conclusion 1645.9. Acknowledgment 1645.10. Bibliography 164Chapter 6. Fuzzy Preference Queries to NoSQL Graph Databases 167Arnaud CASTELLTORT, Anne LAURENT, Olivier PIVERT, Olfa SLAMA and Virginie THION6.1. Introduction 1676.2. Preliminary statements 1686.2.1. Graph databases 1686.2.2. Fuzzy set theory 1746.3. Fuzzy preference queries over graph databases 1766.3.1. Fuzzy preference queries over crisp graph databases 1766.3.2. Fuzzy preference queries over fuzzy graph databases 1826.4. Implementation challenges 1936.4.1. Modeling fuzzy databases 1936.4.2. Evaluation of queries with fuzzy preferences 1936.4.3. Scalability 1956.5. Related work 1976.6. Conclusion and perspectives 1986.7. Acknowledgment 1996.8. Bibliography 199Chapter 7. Relevant Filtering in a Distributed Content-based Publish/Subscribe System 203Cédric DU MOUZA and Nicolas TRAVERS7.1. Introduction 2037.2. Related work: novelty and diversity filtering 2057.3. A Publish/Subscribe data model 2067.3.1. Data model 2067.3.2. Weighting terms in textual data flows 2077.4. Publish/Subscribe relevance 2087.4.1. Items and histories 2087.4.2. Novelty 2097.4.3. Diversity 2097.4.4. An overview of the filtering process 2107.4.5. Choices of relevance 2107.5. Real-time integration of novelty and diversity 2127.5.1. Centralized implementation 2127.5.2. Distributed filtering 2167.6. TDV updates 2217.6.1. TDV computation techniques 2217.6.2. Incremental approach 2237.6.3. TDV in a distributed environment 2257.7. Experiments 2287.7.1. Implementation and description of datasets 2297.7.2. TDV updates 2297.7.3. Filtering rate 2307.7.4. Performance evaluation in the centralized environment 2347.7.5. Performance evaluation in a distributed environment 2387.7.6. Quality of filtering 2407.8. Conclusion 2417.9. Bibliography 242List of Authors 245Index 247

Olivier Pivert is currently a full Professor of Computer Science at the National School of Applied Sciences and Technology, Lannion, France; and a Member of the Institute for Research in Computer Science and Random Systems where he heads the Shaman research team.

• Foreword xiAnne LAURENT and Dominique LAURENTPreface xiiiOlivier PIVERTChapter 1. NoSQL Languages and Systems 1Kim NGUYỄN1.1. Introduction 11.1.1. The rise of NoSQL systems and languages 11.1.2. Overview of NoSQL concepts 41.1.3. Current trends of French research in NoSQL languages 61.2. Join implementations on top of MapReduce 71.3. Models for NoSQL languages and systems 121.4. New challenges for database research 161.5. Bibliography 18Chapter 2. Distributed SPARQL Query Processing: A Case Study with Apache Spark 21Bernd AMANN, Olivier CURÉ and Hubert NAACKE2.1. Introduction 212.2. RDF and SPARQL 222.2.1. RDF framework and data model 222.2.2. SPARQL query language 252.3. SPARQL query processing 292.3.1. SPARQL with and without RDF/S entailment 292.3.2. Query optimization 302.3.3. Triple store systems 332.4. SPARQL and MapReduce 342.4.1. MapReduce-based SPARQL processing 352.4.2. Related work 392.5. SPARQL on Apache Spark 412.5.1. Apache Spark 412.5.2. SPARQL on Spark 422.5.3. Experimental evaluation 482.6. Bibliography 53Chapter 3. Doing Web Data: from Dataset Recommendation to Data Linking 57Manel ACHICHI, Mohamed BEN ELLEFI, Zohra BELLAHSENE and Konstantin TODOROV3.1. Introduction 573.1.1. The Semantic Web vision 573.1.2. Linked data life cycles 583.1.3. Chapter overview 613.2. Datasets recommendation for data linking 623.2.1. Process definition 633.2.2. Dataset recommendation for data linking based on a Semantic Web index 643.2.3. Dataset recommendation for data linking based on social networks 643.2.4. Dataset recommendation for data linking based on domain-specific keywords 653.2.5. Dataset recommendation for data linking based on topic modeling 653.2.6. Dataset recommendation for data linking based on topic profiles 663.2.7. Dataset recommendation for data linking based on intensional profiling 673.2.8. Discussion on dataset recommendation approaches 683.3. Challenges of linking data 693.3.1. Value dimension 703.3.2. Ontological dimension 743.3.3. Logical dimension 773.4. Techniques applied to the data linking process 783.4.1. Data linking techniques 793.4.2. Discussion 833.5. Conclusion 863.6. Bibliography 87Chapter 4. Big Data Integration in Cloud Environments: Requirements, Solutions and Challenges 93Rami SELLAMI and Bruno DEFUDE4.1. Introduction 934.2. Big Data integration requirements in Cloud environments 964.3. Automatic data store selection and discovery 994.3.1. Introduction 994.3.2. Model-based approaches 994.3.3. Matching-oriented approaches 1004.3.4. Comparison 1024.4. Unique access for all data stores 1034.4.1. Introduction 1034.4.2. ODBAPI: A unified REST API for relational and NoSQL data stores 1044.4.3. Other works 1054.4.4. Comparison 1074.5. Unified data model and query languages 1084.5.1. Introduction 1084.5.2. Data models of classical data integration approaches 1094.5.3. A global schema to unify the view over relational and NoSQL data stores 1104.5.4. Other works 1134.5.5. Comparison 1174.6. Query processing and optimization 1184.6.1. Introduction 1184.6.2. Federated query language approaches 1184.6.3. Integrated query language approaches 1214.6.4. Comparison 1244.7. Summary and open issues 1254.7.1. Summary 1254.7.2. Open issues 1274.8. Conclusion 1294.9. Bibliography 129Chapter 5. Querying RDF Data: A Multigraph-based Approach 135Vijay INGALALLI, Dino IENCO and Pascal PONCELET5.1. Introduction 1355.2. Related work 1375.3. Background and preliminaries 1375.3.1. RDF data 1385.3.2. SPARQL query 1405.3.3. SPARQL querying by adopting multigraph homomorphism 1425.4. AMBER: A SPARQL querying engine 1435.5. Index construction 1445.5.1. Attribute index 1445.5.2. Vertex signature index 1455.5.3. Vertex neighborhood index 1485.6. Query matching procedure 1495.6.1. Vertex-level processing 1515.6.2. Processing satellite vertices 1525.6.3. Arbitrary query processing 1545.7. Experimental analysis 1595.7.1. Experimental setup 1595.7.2. Workload generation 1605.7.3. Comparison with RDF engines 1615.8. Conclusion 1645.9. Acknowledgment 1645.10. Bibliography 164Chapter 6. Fuzzy Preference Queries to NoSQL Graph Databases 167Arnaud CASTELLTORT, Anne LAURENT, Olivier PIVERT, Olfa SLAMA and Virginie THION6.1. Introduction 1676.2. Preliminary statements 1686.2.1. Graph databases 1686.2.2. Fuzzy set theory 1746.3. Fuzzy preference queries over graph databases 1766.3.1. Fuzzy preference queries over crisp graph databases 1766.3.2. Fuzzy preference queries over fuzzy graph databases 1826.4. Implementation challenges 1936.4.1. Modeling fuzzy databases 1936.4.2. Evaluation of queries with fuzzy preferences 1936.4.3. Scalability 1956.5. Related work 1976.6. Conclusion and perspectives 1986.7. Acknowledgment 1996.8. Bibliography 199Chapter 7. Relevant Filtering in a Distributed Content-based Publish/Subscribe System 203Cédric DU MOUZA and Nicolas TRAVERS7.1. Introduction 2037.2. Related work: novelty and diversity filtering 2057.3. A Publish/Subscribe data model 2067.3.1. Data model 2067.3.2. Weighting terms in textual data flows 2077.4. Publish/Subscribe relevance 2087.4.1. Items and histories 2087.4.2. Novelty 2097.4.3. Diversity 2097.4.4. An overview of the filtering process 2107.4.5. Choices of relevance 2107.5. Real-time integration of novelty and diversity 2127.5.1. Centralized implementation 2127.5.2. Distributed filtering 2167.6. TDV updates 2217.6.1. TDV computation techniques 2217.6.2. Incremental approach 2237.6.3. TDV in a distributed environment 2257.7. Experiments 2287.7.1. Implementation and description of datasets 2297.7.2. TDV updates 2297.7.3. Filtering rate 2307.7.4. Performance evaluation in the centralized environment 2347.7.5. Performance evaluation in a distributed environment 2387.7.6. Quality of filtering 2407.8. Conclusion 2417.9. Bibliography 242List of Authors 245Index 247

Olivier Pivert is currently a full Professor of Computer Science at the National School of Applied Sciences and Technology, Lannion, France; and a Member of the Institute for Research in Computer Science and Random Systems where he heads the Shaman research team.

• Foreword xiAnne LAURENT and Dominique LAURENTPreface xiiiOlivier PIVERTChapter 1. NoSQL Languages and Systems 1Kim NGUYỄN1.1. Introduction 11.1.1. The rise of NoSQL systems and languages 11.1.2. Overview of NoSQL concepts 41.1.3. Current trends of French research in NoSQL languages 61.2. Join implementations on top of MapReduce 71.3. Models for NoSQL languages and systems 121.4. New challenges for database research 161.5. Bibliography 18Chapter 2. Distributed SPARQL Query Processing: A Case Study with Apache Spark 21Bernd AMANN, Olivier CURÉ and Hubert NAACKE2.1. Introduction 212.2. RDF and SPARQL 222.2.1. RDF framework and data model 222.2.2. SPARQL query language 252.3. SPARQL query processing 292.3.1. SPARQL with and without RDF/S entailment 292.3.2. Query optimization 302.3.3. Triple store systems 332.4. SPARQL and MapReduce 342.4.1. MapReduce-based SPARQL processing 352.4.2. Related work 392.5. SPARQL on Apache Spark 412.5.1. Apache Spark 412.5.2. SPARQL on Spark 422.5.3. Experimental evaluation 482.6. Bibliography 53Chapter 3. Doing Web Data: from Dataset Recommendation to Data Linking 57Manel ACHICHI, Mohamed BEN ELLEFI, Zohra BELLAHSENE and Konstantin TODOROV3.1. Introduction 573.1.1. The Semantic Web vision 573.1.2. Linked data life cycles 583.1.3. Chapter overview 613.2. Datasets recommendation for data linking 623.2.1. Process definition 633.2.2. Dataset recommendation for data linking based on a Semantic Web index 643.2.3. Dataset recommendation for data linking based on social networks 643.2.4. Dataset recommendation for data linking based on domain-specific keywords 653.2.5. Dataset recommendation for data linking based on topic modeling 653.2.6. Dataset recommendation for data linking based on topic profiles 663.2.7. Dataset recommendation for data linking based on intensional profiling 673.2.8. Discussion on dataset recommendation approaches 683.3. Challenges of linking data 693.3.1. Value dimension 703.3.2. Ontological dimension 743.3.3. Logical dimension 773.4. Techniques applied to the data linking process 783.4.1. Data linking techniques 793.4.2. Discussion 833.5. Conclusion 863.6. Bibliography 87Chapter 4. Big Data Integration in Cloud Environments: Requirements, Solutions and Challenges 93Rami SELLAMI and Bruno DEFUDE4.1. Introduction 934.2. Big Data integration requirements in Cloud environments 964.3. Automatic data store selection and discovery 994.3.1. Introduction 994.3.2. Model-based approaches 994.3.3. Matching-oriented approaches 1004.3.4. Comparison 1024.4. Unique access for all data stores 1034.4.1. Introduction 1034.4.2. ODBAPI: A unified REST API for relational and NoSQL data stores 1044.4.3. Other works 1054.4.4. Comparison 1074.5. Unified data model and query languages 1084.5.1. Introduction 1084.5.2. Data models of classical data integration approaches 1094.5.3. A global schema to unify the view over relational and NoSQL data stores 1104.5.4. Other works 1134.5.5. Comparison 1174.6. Query processing and optimization 1184.6.1. Introduction 1184.6.2. Federated query language approaches 1184.6.3. Integrated query language approaches 1214.6.4. Comparison 1244.7. Summary and open issues 1254.7.1. Summary 1254.7.2. Open issues 1274.8. Conclusion 1294.9. Bibliography 129Chapter 5. Querying RDF Data: A Multigraph-based Approach 135Vijay INGALALLI, Dino IENCO and Pascal PONCELET5.1. Introduction 1355.2. Related work 1375.3. Background and preliminaries 1375.3.1. RDF data 1385.3.2. SPARQL query 1405.3.3. SPARQL querying by adopting multigraph homomorphism 1425.4. AMBER: A SPARQL querying engine 1435.5. Index construction 1445.5.1. Attribute index 1445.5.2. Vertex signature index 1455.5.3. Vertex neighborhood index 1485.6. Query matching procedure 1495.6.1. Vertex-level processing 1515.6.2. Processing satellite vertices 1525.6.3. Arbitrary query processing 1545.7. Experimental analysis 1595.7.1. Experimental setup 1595.7.2. Workload generation 1605.7.3. Comparison with RDF engines 1615.8. Conclusion 1645.9. Acknowledgment 1645.10. Bibliography 164Chapter 6. Fuzzy Preference Queries to NoSQL Graph Databases 167Arnaud CASTELLTORT, Anne LAURENT, Olivier PIVERT, Olfa SLAMA and Virginie THION6.1. Introduction 1676.2. Preliminary statements 1686.2.1. Graph databases 1686.2.2. Fuzzy set theory 1746.3. Fuzzy preference queries over graph databases 1766.3.1. Fuzzy preference queries over crisp graph databases 1766.3.2. Fuzzy preference queries over fuzzy graph databases 1826.4. Implementation challenges 1936.4.1. Modeling fuzzy databases 1936.4.2. Evaluation of queries with fuzzy preferences 1936.4.3. Scalability 1956.5. Related work 1976.6. Conclusion and perspectives 1986.7. Acknowledgment 1996.8. Bibliography 199Chapter 7. Relevant Filtering in a Distributed Content-based Publish/Subscribe System 203Cédric DU MOUZA and Nicolas TRAVERS7.1. Introduction 2037.2. Related work: novelty and diversity filtering 2057.3. A Publish/Subscribe data model 2067.3.1. Data model 2067.3.2. Weighting terms in textual data flows 2077.4. Publish/Subscribe relevance 2087.4.1. Items and histories 2087.4.2. Novelty 2097.4.3. Diversity 2097.4.4. An overview of the filtering process 2107.4.5. Choices of relevance 2107.5. Real-time integration of novelty and diversity 2127.5.1. Centralized implementation 2127.5.2. Distributed filtering 2167.6. TDV updates 2217.6.1. TDV computation techniques 2217.6.2. Incremental approach 2237.6.3. TDV in a distributed environment 2257.7. Experiments 2287.7.1. Implementation and description of datasets 2297.7.2. TDV updates 2297.7.3. Filtering rate 2307.7.4. Performance evaluation in the centralized environment 2347.7.5. Performance evaluation in a distributed environment 2387.7.6. Quality of filtering 2407.8. Conclusion 2417.9. Bibliography 242List of Authors 245Index 247