A Natural Language Processing Model to Improve the Software Testing Process Under an Agile Methodology

When verifying the quality of a software product, software testing is the most important phase of the software development life cycle. In 2020, the cost of poor software quality was estimated to be $2.08 trillion in the United States. When using an agile methodology, the software development life cycle is short and fast. When there are any cuts in a project to reduce or maintain the budget, software testing usually is the first step to be impacted, which can cause issues in both the budget and the quality of the software.This praxis presents a machine learning model by developing a natural processing algorithm to predict the classification of the required type of testing for each agile software requirement user story. The software project data set was obtained from a large American medical device company. Five hundred user stories were selected for this praxis from a software development team utilizing the agile methodology.This praxis demonstrates how the term frequency by inverse document frequencyalgorithm can effectively and efficiently improve the software testing process. The algorithm achieves this by processing the number of times a word appears in a user story requirement divided by the total number of words in the user story and calculating the weight of rare words across all datasets. This research establishes a benchmark for the software industry to measure software testing time and cost and to improve methods in the early stages of the software development life cycle.

Author

• Farahanchi, Hamed

Language

• en

Date created

• 2022

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

Degree

• D.Engr.

Advisor

• Yassan, Rebecca

Committee Member(s)

• Campos-Nanez, Enrique
• Dimitriou, Ari

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/z890rv14r