EVALUATION OF AN ALGORITHM OF SOFTWARE DEFECTS OF UNDERSTANDABILITY USING A NEW METRICS OF SOFTWARE

Computer Engineering

Prudence is one of the important features of software quality, as it can affect the stability of software. The cost and reuse of software is also likely to make sense. To maintain software, programmers have to understand the source code. The understanding of source code depends on the psychological complexity of the software, and cognitive abilities are required to understand the source code. The understanding of source code is influenced by so many factors, here we have taken various factors from a unified view. In it we have chosen a rough set approach to calculate comprehensibility based on external identity. Outliers generally have an unusual behavior, here we have taken that the project can be easily understood or difficult to understand. Here we have taken a few factors that underlie understanding, bringing forward an integrated approach to determining understanding. We extracted 20 test case metrics, six developer-related metrics, and two understandable proxies from a white-box test case classification experiment. Based on these matrices, we employed classification and regression algorithms to construct the model to understand the test case. From the experiment, we can conclude that the combined matrices always exhibit better discriminated performance in the classification model as well as a higher correlation in the regression model when compared to a model that only includes test case metrics or developer metrics.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.