A fuzzy set approach to software reliability modeling

This chapter provides a discussion of a fuzzy set approach which is used to extend the notion of software debugging from a 0-1 (perfect/imperfect) crisp approach to one which incorporates some fuzzy sets ideas. The main objective of this extension is to make current software reliability models more realistic. The theory underlying this approach, and hence its key modeling tool, is the theory of random point processes with fuzzy marks. The relevance of this theory to software debugging arises from the fact that it incorporates the randomness due to the locations of the software faults and the fuzziness bestowed by the imprecision of the debugging effort. Through several examples, we also demonstrates that this theory provides the natural vehicle for an investigation into the properties and efficacy of fuzzy debugging of software programs and is therefore a contribution to computational intelligence