Title
Improving source selection in large scale mediation systems through combinatorial optimization techniques
Abstract
This paper concerns querying in large scale virtual organizations. Such organizations are characterized by a challenging data context involving a large number of distributed data sources with strong heterogeneity and uncontrolled data overlapping. In that context, data source selection during query evaluation is particularly important and complex. To cope with this task, we propose OptiSource, an original strategy for source selection using combinatorial optimization techniques combined to organizational knowledge of the virtual organization. Experiment numerical results show that OptiSource is a robust strategy that improves the precision and the recall of the source selection process. This paper presents the data and knowledge models, the definition of OptiSource, the related mathematical model, the prototype and an extensive experimental study.
Year
DOI
Venue
2011
10.1007/978-3-642-23074-5_6
T. Large-Scale Data- and Knowledge-Centered Systems
Keywords
Field
DocType
improving source selection,challenging data context,knowledge model,large number,data source selection,combinatorial optimization technique,source selection,virtual organization,large scale mediation system,large scale,uncontrolled data,data source,source selection process,combinatorial optimization
Data source,Data mining,Computer science,Combinatorial optimization,Mediation (Marxist theory and media studies),Artificial intelligence,Recall,Data, context and interaction,Machine learning,Virtual organization
Journal
Volume
Citations 
PageRank 
3
1
0.35
References 
Authors
24
4
Name
Order
Citations
PageRank
Alexandra Pomares1108.69
Claudia Roncancio224332.92
Van-Dat Cung3799.01
María-del-Pilar Villamil431.73