Semantic analysis of natural language software requirement

In the recent past, domain specific solutions for detailed semantic analysis have got acceptable by natural language processing community and use of applications involving natural language based user interface. Different approaches that has been previously used is focusing on quality of text and improving the text contents by adding semantic information with text then the existing approaches used for semantic analysis can provide better results. In this, an approach was presented to address the problem of non-availability of semantic information required for better semantic analysis. This problem is solved by using semantic technology to annotate text of software requirements expressed in a natural language with their domain specific semantics and investigate the effect of semantic analysis with attached semantics. The presented approach uses a semantic framework specifically designed for interpretation and detailed semantic analysis of natural language software requirement specifications. The used framework is based on semantic technology involves knowledge extracted from existing software requirement documents and knowledge extracted from existing applications. The presented approach shows that by adapting and combing existing ontologies to support knowledge management, developing system and performing experiments on requirement of real world software systems. In this approach start with software requirement specification, after this clean the irrelevant requirements, convert the cleaned requirements into graph that represents inter related different elements. Represent the requirement graph into sparse matrix, after these all steps; we generate ontology with the help of OntoGen tool.