A Multi-label Marking Method for Defect Reports

Abstract

The invention provides a multi-label marking method for defect reports. In the present invention, the user specifies a set of labels to be selected, constructs an initial multi-label training data set according to the set, performs the first round of label completion through a manual completion method to obtain a multi-label training data set after the first round of completion, and uses binary classification The prediction method automatically completes the second round of label completion to obtain the multi-label training data set after the second round of completion, and then performs the third round of label completion through the manual completion method to obtain the final multi-label training data set; build a fast text classifier based on The text multi-category classification model of the algorithm is trained based on the final multi-label training data set, and the trained text multi-classification model is used as the post-training defect report label predictor; use the post-training defect report label predictor to perform multi-label labeling for defect reports , and collect user feedback to update prediction performance on the fly. The present invention provides a method for efficiently predicting and labeling multiple labels for defect reports.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a multi-label marking method for defect reports. Background technique [0002] Defect reports play an important role in software development. The increasingly mature defect tracking system, along with the continuous improvement of defect report management capabilities, also provides more convenient defect report sorting services for modern software development. For example, GitHub, an all-in-one coding community, has integrated a Git-based version control system with a lightweight defect tracking system to provide convenient coding services. This bug report tracking system helps to log thousands of bug reports for many open source projects. [0003] A very important function of a bug report tracking system is to assign tags to bug reports. The tags of defect reports help developers to quickly classify or identify defect reports. [0004] In real projects, defect reports ofte...

AI Technical Summary

Problems solved by technology

However, if you want to add labels to defect reports from cross-warehouses, how to filter and unify labels is a problem that label classification will encounter

The existing single-label methods have some work in the unification of labels, but the types of unified labels are not enough to meet the data requirements of multi-label

Images