Abstract
As an important part of software engineering, software testing is a knowledge-intensive work. In the process of software testing, inconsistent knowledge expression, diverse knowledge carriers, and a few experienced people have mastered most of the knowledge, which hinders the transfer and sharing of domain knowledge. Ontology is widely used in various stages of software engineering to define the semantic relationship between relevant information and knowledge. To solve the problem of knowledge silo in the process of software testing, this article forms an Entire Process Ontology on Software Testing (EPOST). EPOST covers the concepts and relationships of software testing process information, software test object information, and software defect information. The concepts and terms in the ontology are extracted from ISTQB, SWEBOK, IEEE std.829-2008 standard, and IEEE std.610.12-1990 standard. We adopt a comprehensive ontology construction method based on Dev. 101 method and Methontology method. The developed ontology is successfully evaluated by using validation and verification tests. Ontology verification uses an improved FOCA evaluation method by adding a cohesion metric. The evaluation result infers that EPOST has a high quality of ontology and good domain coverage, and achieves the purpose of ontology construction. Finally, we make a case study on the role of EPOST in software testing process. The results show that ontology-based application in the software testing process can promote the sharing and transmission of domain knowledge, and improve the testing process and testing quality.
Highlights
Software testing is an important means of software quality assurance, and most of the cost of software development and maintenance is test cost [1]
The proportion of problems detected by using Entire Process Ontology on Software Testing (EPOST) is higher, indicating that the size of knowledge affected the quality of testing results
The evaluation method is the core of ontology evaluation, the evaluation metric system is the basis of ontology evaluation, and the evaluation tool is the realization method of ontology evaluation
Summary
Software testing is an important means of software quality assurance, and most of the cost of software development and maintenance is test cost [1]. Software testers need to master a large number of testing technologies and tools and need to understand the business characteristics of the tested objects, which makes software testing knowledge-intensive work. Given the different personal experiences and abilities of testers, it is prone to incomplete consideration of abnormal situations during software testing requirements analysis, which leads to the problem of missing items in the test criteria and test case design. In terms of use case design, tool selection, report preparation, etc., there is no unified process and guidelines to guide, it is difficult to categorize and reuse, resulting in the uneven quality of test results. To effectively avoid the problem of knowledge silo in the software testing domain due to the large, scattered, and difficult to obtain test data, it is VOLUME XX, 2020
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
