Enriching SPARQL Queries by User Preferences for Results Adaptation

Abstract

Systems of data integration using ontologies aim to implement a collaborative environment between sources for sharing data and services to respond a user request for information. Their users’ requests are an exact expression of their needs. However, the multiplicity of data sources, their scalability and the increasing difficulty to control their descriptions and their contents are the reasons behind the implacability of this assumption today. The users now may not know the data sources they questioned, nor their description or content. Consequently, their queries reflect no more a need that must be satisfied but an intention that must be refined according to data sources available at the time of interrogation. In this work, we present a semantic-based approach to enrich user’ queries expressed in SPARQL Language by his preferences in order to adapt the returned results and make them more precise and more relevant. The proposed approach is applied on a movies management system based on the standard MovieLens dataset. The obtained results are compared to existing approaches according to precision and recall measures. Our approach improved the precision with 26% and the recall with 7% comparing to those of previous study using collaborative filtering.