ISBN: 978-981-11-3671-9 DOI: 10.18178/wcse.2017.06.040
Software Requirement Feature Selection Based On Tabu Search Algorithm
Abstract— Software requirement classification plays a key role in test case reuse, and the most important
part for requirement classification is the feature selection algorithm. In the classification process, the quality
of the feature that been selected can seriously impact on the result of classification. In order to improve the
feature selection, the tabu search had been used to apply different requirement feature combinations, which
can effectively avoid getting in the local best solution, and finally obtain the optimal attribute subsets. In
experiment, Chinese Corpus had been used to verify our algorithm and strategy, and results showed that,
comparing with genetic algorithm, our method can effectively remove the invalid characters and improve
classification accuracy, which means this method can ensure the accuracy of software requirement
classification.
Index Terms— software requirement classification, feature selection, tabu search
Tong Wang, Ying Shang
College of Information Science and Technology, Beijing University of Chemical Technology, CHINA
Cite: Tong Wang, Ying Shang, "Software Requirement Feature Selection Based On Tabu Search Algorithm," Proceedings of 2017 the 7th International Workshop on Computer Science and Engineering, pp. 233-238, Beijing, 25-27 June, 2017.