From User Requirements to Querying of Fuzzy Summaries

Abstract

With the huge amount and the evolution of fuzzy data, the necessity to work with synthetic views became a challenge for many databases (DB) community researchers. Data summarization techniques are now considered as accurate tools to handle huge DB, in particular when precise data are not needed. Formal approaches have been proposed making possible the generation of an hierarchy of summaries from DB. The challenges arise on the question of how querying these fuzzy views according user requirements. In this work, we propose to handle with these challenges by query repairing and substitution. Two process were studied, the first process is used by modifying query while using the best fuzzy summaries which have the most near answers. The second one is applied to generate all substitution queries over the fuzzy summaries' hierarchy. This would be not only expensive but also unjustified for the part of the search hierarchy nodes.