Software Metric for Assessing the Quality of Software Requirements

Elly, Stephen (2020) Software Metric for Assessing the Quality of Software Requirements. Masters thesis, Universiti Malaysia Sarawak (UNIMAS).

	PDF Elly - 24.pdf Download (731kB)
	PDF (Please get the password by email to repository@unimas.my, or call ext: 3914/ 3942/ 3933) Elly anak Stephen ft.pdf Restricted to Registered users only Download (1MB) | Request a copy

Abstract

SRS document composes of FR and NFR. However, due to the heterogeneous domain development environment, the quality of the produced SRS document is in question. Software development depends on the quality of the SRS document. Poor quality of the SRS document results in poor software production. A study on the sample of the SRS document shows a lack of standardization of the document structure as well as its FR. A recent researcher manually evaluates the SRS document which is time consuming. Furthermore, the unstructured way in writing FR leads to ambiguity. A recent researcher shows that the ambiguity of natural language can be reduced by restricting the term used in writing FR. Numbers properties for each quality had been introduced to evaluate the written FR. Two methods are adopted in order to standardize the structural and FR. The first method is by comparison with the IEEE 830 ToC. The topic from the ToC is extracted and used to assess the structure of the document. The second method is the RB. It is used to standardize the way of defining the FR. The structural and FR are assessed based on four quality properties which are completeness, consistency, correctness and preciseness. The completeness quality is intended to assess the SRS structure meanwhile the others are for FR. A framework is designed based on quality properties to be assessed. The quality properties are formalized by proposing the equation for measurement. The framework is implemented and the prototype is tested to evaluate its capability. Case study is applied to the prototype. This resulted in two outputs. The first output is by comparison against the IEEE 830 which results in the percentage similarity of structure. Meanwhile, second output resulted from prototype analyst the user input. The overall measurement of the case study is produced by a prototype in the degree of percentage by adding both from the first and second outputs. From the results, the developer can evaluate the maturity of the SRS document. The resulting measurement of the SRS structure allows the developer to reorganize the document to increase readability. The result from the measurement of FR allows the developer to ensure the consistency, validating and avoiding unspecific usage of the term used in writing the FR.

Actions (For repository members only: login required)

View Item