F. Javier Crespo, Oscar Marbán. On the Use of Tools Based on Fuzzy Set Theories in Parametric Software Cost Estimation. In Vicenç Torra, Yasuo Narukawa, Aïda Valls, Josep Domingo-Ferrer, editors, Modeling Decisions for Artificial Intelligence, Third International Conference, MDAI 2006, Tarragona, Spain, April 3-5, 2006, Proceedings. Volume 3885 of Lecture Notes in Computer Science, pages 129-137, Springer, 2006. [doi]
The whole software industry has an awful footpath for delivering software on-time and on-budget. Probably, one reason is coming from not deal with the imperfection of information when they use a lot of human process. In this paper, we propose the use of fuzzy measures in contrast with crisp measures of traditional models and, therefore, apply of appropriate aggregators. Traditional models of software cost estimation are constructed from project databases and they describe cost drivers in terms of linguistic estimations using vague terms like “low” or “high”, and such expressions are also used in obtaining actual predictions. But cost drivers are in many cases abstract concepts that are better estimated by breaking them down in a number of second-level aspects. The method proposed is based both, on a concrete study of the use of linguistic variable human categorizations and, on level aspects that are defined by layer and are easy to raise using appropriate aggregators. Moreover, the proposed scheme can have different planes according to the model morphology.