Relational data warehouses and aggregation

For more info about our services contact : help@bestpfe.com

Table of contents

1 introduction 
1.1 Motivation and outline
1.2 Manuscript organization
2 preliminaries 
2.1 Data graphs and the Resource Description Framework
2.2 Queries on RDF graphs
2.3 Relational data warehouses and aggregation
3 rdf graph summarization 
3.1 Motivation
3.2 Background: quotient RDF graph summarization
3.2.1 Quotient graphs
3.2.2 Prior work
3.2.3 Limitations of the prior work
3.2.4 Notation
3.3 Data graph summarization
3.3.1 Data property cliques
3.3.2 Strong and weak node equivalences
3.3.3 Weak and strong summarization
3.4 Typed data graph summarization
3.4.1 Data-then-type summarization
3.4.2 Type-then-data summarization
3.5 Summarization of graphs with RDFS ontologies
3.5.1 Type-then-data summarization using most general types
3.5.2 Interactions between summarization and saturation
3.5.3 Shortcut results
3.5.4 Relationships between summaries
3.6 From summaries to visualizations
3.6.1 Leaf and type inlining
3.6.2 Summary statistics
3.6.3 Visualizing very large summaries
3.7 Summarization algorithms
3.7.1 Global data graph summarization
3.7.2 Incremental data graph summarization
3.7.3 Global and incremental typed graph summarization
3.7.4 Parallel summarization
3.7.5 Parallel data graph summarization
3.7.6 Parallel typed graph summarization
3.7.7 Apache Spark implementation specifics
3.8 Centralized summarization experiments
3.8.1 Centralized algorithms compared
3.8.2 Datasets
3.8.3 Summary size
3.8.4 Summarization time
3.8.5 Summary precision
3.8.6 Experimental conclusions
3.9 Parallel summarization experiments
3.9.1 Configuration
3.9.2 Speedup through parallelism
3.9.3 Scalability study
3.9.4 Experimental conclusions
3.10 Non-quotient RDF graph summarization
3.11 Conclusion
4 discovering interesting aggregates in rdf graphs 
4.1 Motivation
4.2 Problem statement and notation
4.3 Overview of the approach
4.4 Lattice-based computation
4.4.1 Incorrectness in the RDF setting
4.4.2 MVDCube algorithm
4.5 Early-stop aggregate pruning
4.5.1 Early-stop principle
4.5.2 Estimating the interestingness score
4.5.3 Other interestingness functions
4.5.4 Plugging early-stop into MVDCube
4.6 Implementation details
4.6.1 Offline processing
4.6.2 Online processing
4.7 Experimental evaluation
4.7.1 Analysis of example results
4.7.2 Benefits of derived properties
4.7.3 Analysis of MVDCube against PGCube
4.7.4 Impact of early-stop on MVDCube
4.7.5 Scalability study
4.7.6 Experimental conclusions
4.8 Related work
4.9 Conclusion
5 conclusion and perspectives 
5.1 Thesis conclusion
5.2 Future work perspectives
bibliography