Abstract
Software fault-prone prediction is one of the active areas of software engineering. It plays a very important role in the analysis of software quality and balance of software cost. Practically, the identification of a module's fault-prone is very important for minimizing cost and improving the effectiveness of the software development process. Software fault-prone prediction helps us to develop dependable software. How to obtain the correlation between software metrics and module's fault-prone, hiding in the observed metrics data, has been focused by many researches. In this paper, we propose the use of a fuzzy integral (FI) for this purpose. FI offers significant advantages over other approaches due to its ability to naturally represent qualitative characteristic of software fault-prone. Proposed approach was applied on Chidamber-Kemerer (CK) metrics and two datasets of NASA Metrics Data Program from PROMISE repository. Experiments results confirm that proposed approach is very effective for establishing relationship between software metrics and fault-prone. Its implementation doesn't require expert's knowledge. Proposed approach can give useful results for software project managers.