Method for program error location based on conditional probability

Abstract

The invention discloses a method for program error location based on conditional probability. According to the method, by empirical research on inherent relevancy between a program frequency spectrum and execution results, conditional probability thought of statistics is introduced, and a P model for quantitative analysis on relationship strength evaluation of the program frequency spectrum and execution results is constructed. According to some classical SFL methods, a specific anomaly degree calculation formula is analyzed by means of the P model put forward, and thus essential reasons for defect location precision and efficiency difference are studied. In order to verify effectiveness of the method, in empirical research, 7 programs in a Siemens suite and a Space program are taken as standard evaluating objects, and test results indicate that compared with 13 existing classical defect location methods, the method has a better defect location effect.

Description

technical field [0001] The invention belongs to the technical field of software program error location detection, and in particular relates to a program error location method based on conditional probability. Background technique [0002] The increasing scale and complexity of software has brought great challenges to software development and debugging technology. In the face of software defects that follow in the software development process, software testing is an important technical means to improve its quality and reliability. Therefore, people's investment in software testing is increasing year by year. During the software testing process, when the behavior of the tested software is inconsistent with expectations (that is, failure), developers need to carry out software debugging immediately. The primary task of software debugging is the defect Positioning, which provides the basis for subsequent bug fixes. [0003] Defect localization aims to detect and find the wrong ...

AI Technical Summary

Problems solved by technology

[0003] Defect localization aims to detect and find faulty codes that cause software failures, and is a very tedious and time-consuming activity

The traditional manual debugging method of setting breakpoints is not only difficult to select the breakpoint location, but also has a huge time cost; therefore, the automation of defect location has become a common goal pursued by software academia and industry

At present, the automatic defect localization method is not yet mature, and it is still a research hotspot in the field of software engineering. In recent years, researchers have tried to propose a series of defect localization methods from different perspectives, including fault localization methods based on slices, and fault localization methods based on invariants. Localization methods, model checking methods, error localization methods based on program spectrum, etc.

Images