Dynamic proportion test case sorting selection method and system based on machine learning

Abstract

The invention discloses a dynamic proportion test case sorting selection method and system based on machine learning, and belongs to the technical field of software testing. Historical version execution results of a program to be tested are collected, training samples are made, the training samples of different programs are used for forming a universal training set, and the universal training set is used for training a machine learning model; test case information of the to-be-tested program is collected, priority calculation is performed on the test case information, and a test case sorting list is generated; a test result is predicted by using a machine learning model, and a dynamic test case selection proportion is calculated; and according to the test case sorting list and the dynamic test case selection proportion, selection execution of the test case set is performed in a descending order.

Description

technical field [0001] The invention is based on the test case sorting technology commonly used in machine learning and regression testing, realizes a dynamic proportional test case selection method, and belongs to the technology in the field of software testing. Background technique [0002] With the development of the computer industry, the application of software in life and production is becoming more and more extensive. Software engineering is usually a team effort. The traditional way of practice is that each member of the team completes their own work separately, and then integrates the code. However, in the process of code integration, the codes of different members may conflict with each other, resulting in program errors. To solve this problem, software development teams started using continuous integration methods. Continuous integration encourages every member of the team to submit and integrate code frequently, in order to find possible errors in the code int...

AI Technical Summary

Problems solved by technology

However, taking the above two representative technologies as examples, the test case ranking selection technology uses the same test ratio for different builds, ignoring the differences between different builds; the build test failure prediction technology has a degree of discrimination for different builds, but its The result of the execution is the dichotomous result

If the machine learning model predicts a false negative result, i.e. misclassifies a test-failing build as a test-successful build, a bug in the program cannot be detected

Images