Flexible query conditions could use linguistic terms described by fuzzy sets. The question is how to properly construct fuzzy sets for each linguistic term and apply an adequate aggregation function. For construction of fuzzy sets, the lowest value, the highest value of attribute and the distribution of data inside its domain are used. The logarithmic transformation of domains appears to be suitable. This way leads to a balanced distribution of tuples over fuzzy sets. In addition, users’ opinions about linguistic terms as well as current content in database are merged. The second investigated issue is selection of an adequate aggregation operator. Usual t-norm functions as well as compensatory γ – operator have been examined. Finally, the interface for managing these issues has been proposed. A user can obtain an overview about stored data before running a query; that may reduce empty or overabundant answers.
Fuzzy Queries; Construction of Fuzzy Sets; Aggregation Operators; Database
Main Research Area:
14th International Conference on Enterprise Information Systems (ICEIS 2012)