Software defect prediction method and system based on TAN semi-naive Bayesian network

Abstract

The invention provides a software defect prediction method and system based on a TAN semi-naive Bayesian network. The method comprises the following steps: collecting historical defect records to form a training data set; summarizing historical items in the training data set; performing word segmentation on a to-be-predicted text to obtain a segmented word set; carrying out merging processing on repeated segmented words in the segmented word set to form a feature word set; calculating a condition mutual information value between any two feature words in the feature word set; constructing an undirected graph by taking each feature word in the feature word set as a node and the mutual information value as an edge; selecting any node as a root node, taking the outward direction of the root node as the circulation direction between the nodes, sequentially setting the circulation direction between the nodes in the undirected graph set in a recursive mode, and forming a directed acyclic graph; constructing a TAN semi-naive Bayesian network, calculating and comparing the posterior probability value of each classification in the prediction classification set based on the word segmentation set, and taking the classification with the maximum posterior probability value as a final software defect prediction result.

Description

technical field [0001] The present invention relates to the technical field of software defect prediction, in particular to a software defect prediction method and system based on TAN semi-naive Bayesian network. Background technique [0002] The application fields of software are very extensive. In many modern large-scale equipment systems, more and more key functions are realized by software. In a software system with a scale of more than one million lines of code, more than 80% of the functions are realized by software. At the same time, the higher the proportion of software, the more serious the problems caused by software defects. In order to effectively avoid the problems caused by software defects, it is necessary to analyze and summarize software defects with representative characteristics from the software defect data of historical projects, and use intelligent methods such as machine learning to train the sorted out software defect data. Research software require...

AI Technical Summary

Problems solved by technology

However, this patent cannot solve the problems of large space overhead, poor computing performance, and low prediction accuracy of the fully connected Bayesian network classification prediction model.

Images