Abstract:
Advances in computer technology in support of further software development have been a factor resulting in the creation of many software engineering tools, for use in the successful development of the system at each step of software development. In each step, it is necessary to use different software engineering tools based on features and usage functions. Also, the software engineering tools are stored across various organizations and websites, making retrieval for usage difficulty.This thesis has solved the aforementioned problems by developing a tool for classifying and retrieving software engineering tools to produce appropriate results for target usage, by using SWEBOK (Software Engineering Body of Knowledge), a software engineering standards document, to assist in the classification of various software engineering tools. The sub-categories of software engineering tools specified by SWEBOK are used as the classifiers. Therefore, the system consists of 26 types of tools. A Naive Text Classification algorithm is used for classifying the various software engineering tools into the 26 categories specified by SWEBOK, and retrieval is based on calculating a similarity value using the Vector Space Model.The experiment to measure the effectiveness of the research was conducted, by comparing the results with the classic information retrieval approach. The effectiveness of software tool retrieval based on the categories specified by SWEBOK is measured using recall and precision values. The results of the experiment show that, the presented method produces a 12.82% lower recall value than the Classic IR Approach, and a 12.73% higher precision value than the Classic IR Approach.