Enhancing the Fuzzy Set Model with Positively Compensatory Operators

The conventional fuzzy set model uses the MIN and MAX operators for the evaluation of boolean operators AND and OR, respectively. It has been criticized to generate incorrect ranked output in many cases. This is because the MIN and MAX operators do not well correspond with human behavior for document ranking. There are a variety of fuzzy operators to replace the MIN and MAX operators. In this paper we first investigate the behavioral aspects of the fuzzy operators and address important issues to affect retrieval effectiveness. We then present the enhanced fuzzy set model based on positively compensatory operators, which are suitable for achieving high retrieval effectiveness. It is also shown through experiments that the enhanced fuzzy set model provides similar retrieval effectiveness and higher retrieval efficiency in comparison with the extended boolean model.