Automatic software traceability recovery method for enhancing data preprocessing process

Abstract

The invention discloses an automatic software traceability recovery method for enhancing a data preprocessing process, which comprises the following steps: selecting a product of which a tracking relationship is to be recovered, extracting related fields of the product to carry out data cleaning and carrying out feature engineering to obtain a sample data set; dividing the sample data set into a marked data set and a missing tracking data set by using a label marking method; segmenting the marked data set into a marked training set and a test set by using a four-fold time sequence verification method; combining the mark training set and the missing tracking data set by using a semi-supervised unbalanced learning framework to generate a new training set; using a plurality of resampling modes, balancing a training set, training a dichotomy model, evaluating the performance of the dichotomy model, and recovering the tracking relation between products. Starting from an enhanced data preprocessing process, the problems of many project products, poor data quality, unbalanced sample data and the like are solved through multiple enhancement measures, and the F1 value, the accuracy rate and the recall rate are greatly improved.

Description

technical field [0001] The invention relates to the technical field of computer technology, in particular to an automatic software traceability restoration method for enhancing the data preprocessing process. Background technique [0002] Software traceability is the ability to relate any uniquely identifiable software artifact to other artifacts, maintain and use the resulting network to answer questions about the software product and its development process. Software traceability technology is dedicated to creating or maintaining the traceability relationship between different artifacts, which helps to improve the quality of process-oriented data. However, software traceability is a difficult and error-prone task. The main difficulty comes from how to fill the logical abstraction gap between the requirements written in natural language and the code written in programming language. At the same time, in the face of the huge number of products and the number of potential tr...

AI Technical Summary

Problems solved by technology

In the face of the huge number of products and the potential tracking relationship between products in enterprise projects, the problem of data sample imbalance caused by the positive and negative labeling strategy in the binary classification model will become more and more serious. At the same time, how to effectively maintain the relationship between products in enterprises There is no long-term stable strategy for the traceability of the tracking relationship, and temporary measures are taken. The above problems make the automatic traceability recovery method of open source projects unable to be directly applied to enterprise projects.

Images