A framework of software requirements quality analysis system using case-based reasoning and Neural Network

In this paper, we propose a new approach to Software Requirements Specifications (SRS) or software requirements quality analysis process. We apply the Software Quality Assurance (SQA) audit technique in determining whether or not the required quality standards within the requirements specifications phase are being followed closely. Quality analysis of the SRS is performed to ensure that the software requirements among others are complete, consistent, correct, modifiable, ranked, traceable, unambiguous, and understandable. Here, a new approach that combines case-based reasoning (CBR) and neural network techniques in analyzing SRS quality is proposed. This approach is used in improving the process of analyzing the quality of a given SRS document for a specific project. The CBR technique is used to evaluate the requirements quality by referring to previously stored software requirements quality analysis cases (past experiences). CBR is an artificial intelligence technique that reasons by remembering previously experienced cases, and this technique will speed up the quality analysis process. Neural Network (Artificial Neural Network or ANN) is the type of information processing paradigm that is inspired by the way biological nervous systems (brain) process information. Neural network technique works well with CBR because it also uses examples to solve problems. The new approach proposed in this research aims at enhancing and improving existing methods in analyzing SRS quality. A framework of the proposed approach is the main outcome of this research study.