Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


44082.pdf (2.44 MB)

ETD Abstract Container

Abstract Header

Clustering abstractions to increase the efficiency of requirements-based testing

Rathod, Prachi Basant
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1659535518422242

Abstract Details

2022, MS, University of Cincinnati, Engineering and Applied Science: Computer Science.
Software systems often fail due to programming errors, human errors, and poor practices implemented during the testing and quality assurance phase. These failures sometimes can be either minuscule or can be life-threatening, such as the Therac-25 radiotherapy incident [1]. We cannot afford both of these errors caused due to inefficiencies in testing practices. The establishment of standard practices and tools developed helped to reduce software failures. These practices are highly recommended to be incorporated during the initial stages of testing. One such test is Requirements-Based Testing (RBT) [2]. Researchers have emphasized enough advantages of the RBT process. They have also developed tools to assist the early phases of Requirements Engineering (RE). RBT helps in requirements validation, analysis, and elicitation. These developments have helped to avoid failures at an earlier stage in the Software Development Life Cycle (SDLC) and reduce the cost of fixing the faults later. One such research focused on the understanding that these requirements reside in the environment[3] of the deployed software. When developers build the intended software with the specific requirements expressed by the stakeholder, it should not violate the environment in which the software deploys. Misinterpretations about the environment because of limited knowledge and vague assumptions can increase the possibility of a breach in said environment’s requirements. In our proposed work, we aim to assist the RBT process by providing a list of abstractions that can capture the essence of environmental assumptions. An abstraction is a term that refers to a particular significance in each domain. These abstractions can help increase the RBT process’s efficiency as they can help reveal bugs in the software. We use relevant Wikipedia pages as our domain corpus and extract abstractions framed by Natural Language Processing (NLP) techniques such as Part-Of-Speech (POS) tagging and Dependency Parsing (DP). Further, we cluster the list of candidates using clustering techniques to detect patterns. We evaluate our approach to six software systems from two application domains. The results show that our approach can assist a human analyst in revealing bugs and also match with the manually created environmental assumptions.
Nan Niu, Ph.D. (Committee Member)
Xuefu Zhou, Ph.D. (Committee Member)
Boyang Wang, Ph.D. (Committee Member)
77 p.
Computer Science
Requirements-based testing; abstractions; natural language; environmental assumptions

Recommended Citations

Citations

  • Rathod, P. B. (2022). Clustering abstractions to increase the efficiency of requirements-based testing [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1659535518422242

    APA Style (7th edition)

  • Rathod, Prachi Basant. Clustering abstractions to increase the efficiency of requirements-based testing. 2022. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1659535518422242.

    MLA Style (8th edition)

  • Rathod, Prachi Basant. "Clustering abstractions to increase the efficiency of requirements-based testing." Master's thesis, University of Cincinnati, 2022. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1659535518422242

    Chicago Manual of Style (17th edition)

ucin1659535518422242
69
© 2022, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.