Extraction of Method Signatures from Ontology Towards Reusability for the Given System Requirement Specification

Software reuse improves productivity, quality, and maintainability of software products. Only few completed projects are achieved and documented. The method signatures in a completed project are stored in the Ontology and the source code components are stored in Hadoop Distributed File System (HDFS). Methods are needed for the new project can be extracted from the Ontology using Software Requirement Specification (SRS) document. UML design document will evolve after many phases from SRS and hence this work proposes a new framework to extract keywords from SRS and estimate the number of new methods to be developed and count the number of methods that can be reused from the Ontology. The SRS document for the project consists of purpose, scope, system requirements, functional requirements and non-functional requirements as metadata. The SRS document is given as input and the keywords are extracted. The keywords are searched in Ontology for the similar method prototypes and the appropriate code components would be extracted from the HDFS. These methods are integrated in the new project with a review process. The implementation is provided with the sample SRS text. The keywords are extracted and matched with the Ontology. The reusability is measured using reuse metrics, quality, and knowledge growth. Index Terms Metadata, Knowledge Management, Ontology, Reusability, WordNet.