Selecting Proper Security Patterns Using Text Classification
Seyed Mohammad Hossein Hasheminejad, Saeed Jalili · 2009
Over the last few years, a large number of security patterns have been published. However, this large number of patterns led to a problem in selecting appropriate patterns for different security requirements. In this paper, we present an automatic selection approach for security patterns. We use text processing approach and learning techniques to select appropriate security patterns for given security problems. In the proposed method, for each security pattern class, one classifier is learned, then proper security patterns are recommended for given security requirement. We also evaluate our proposed method by several security patterns mentioned in Schumacher's Book. As a result of experiments, Naive Bayes is the best learning technique, compatible with the essence of this problem.