Software Requirements Validation : A Comparative Study

As software requirements are the core of building any software project, the requirements engineering became a very important aspect for a lot of researchers. This paper sheds light on the software requirements validation process as an important step of the requirements engineering process that affects the other software production processes such as design, development and testing. The paper compares two different techniques of software requirements validation, each of them is trying to use the user’s language that is the natural lan-guage to enhance the requirements process, while they are different in the modeling of this natural language. The first approach builds conceptual models based on the natural language; these models are easy to be understood by the analyst and the user. The conceptual models can be converted to the ordinary models such as JSD, OCD and DFD, hence, the validation of software requirements will be provided by converting those conceptual models again to the natural language and by finding the gap between the requirements docu-ment and the analysis of the analysts. The second approach builds an assessing tool based on the natural language, this tool tries to find the problems with the requirements document before passing it to the analysis phase, hence, the requirements can be checked again with the user. The comparison takes three basic steps to compare the two approaches, step 1 compares the approaches concepts and tools, step 2 applies a case study to the two approaches and compares the results and step 3 compares the limitations of the two ap-proaches. Finally, the paper reaches a proposed model that combines the two approaches and can improve the software requirements validation process