Data Source Selection Strategy Based on Probability Statistics and Quality Measurement
Zhang Chun · 2004
Data source selection strategy plays an important role in improving the efficiency and effectiveness of heterogeneous data integration. This paper proposes a greedy algorithm to select proper data sources by combining probabilistic knowledge in mediated system and qualitative information about quality of data sources. Two kinds of probabilistic information of frequent topic classes are considered,which are coverage of each data source and overlap degree between different data sources. Quality of data sources includes query price,response time,reliability and user satisfaction. Finally,a preliminary experiment is carried out to evaluate the efficiency of the proposed algorithm.