A Rule Extraction for Outsourced Software Project Risk Classification

A rule extraction algorithm based on K-means clustering with interval-valued intuitionistic fuzzy sets (IVIFS) information, which is the combination of K-means clustering and IVIFS, is proposed in this paper. First, we introduce IVIFS and its distance. Second, we introduce IVIFS method and its application to the classification of software project. Finally, we present a rule extraction model according to IVIFS fuzziness and its K-means clustering, and apply them to pattern classification of outsourced software project risk to demonstrate the advantages of this model. The experimental results show that the rules from IVIFS model are better than that from the conventional K-means clustering model in rule extraction, and the prediction effect from the former is more effective than that from the latter. According to this combinational rule extraction method, based on database from a special and professional investigation for Chinese small and medium software-outsourced enterprise, we obtain some valuable, realistic and available project development risks decision rules.