Multi-source software development data fusion-oriented developer portrait modeling method

Abstract

The invention relates to a multi-source software development data fusion-oriented developer portrait modeling method. According to the method, features and APIs are extracted from README.md and code text in GitHub and code text and description text in Answers in Stack Overflow, similarity degrees of the features and a similarity degree of the APIs are calculated, then comprehensive similarity degrees of the both are calculated and sorted, it is considered that a developer1 of the GitHub and a developer2 of the Stack Overflow are the same user when one of the comprehensive similarity degrees is greater than a certain value (theta), data belonging to the same user are fused, and skills and activities of the developer, relationships between the developer and code and relationships between developers are respectively mined, and are comprehensively shown by using a portrait. The method overcomes a defect that recommendation cannot be provided for fully completing a software engineering task. The method utilizes the multi-source data and combination of the developer and the code to model the portrait of the developer.

Description

technical field [0001] The invention belongs to the field of software engineering, in particular to a developer portrait modeling method for data fusion of multi-source software development. Background technique [0002] There are various types of software development and maintenance tasks in software engineering, such as software development, testing, etc., and the skills and experience of each developer are different, so how to assign reasonable tasks to developers is the key to software development and maintenance. A big problem in the field of maintenance. Modeling the developer’s portrait can solve this problem. However, the developer’s portrait modeling method based on single-source data cannot fully describe the developer’s portrait. This method has the problem of data sparsity, which leads to insufficient skill extraction of developers. Comprehensive, it can only discover the developer's activities in a single website and the explicit relationship between the develo...

AI Technical Summary

Problems solved by technology

However, these technologies can only one-sidedly excavate the professional knowledge of the developer, the position in the same organization, the role in the software development process, and the interaction information with other developers in the same project. There is a singleness in the portrait. In terms of personalized recommendations, developer portraits are used to filter recommended API methods, display software interfaces that developers are interested in, etc. In terms of recommending developers, recommend developers to solve user support request problems and help users solve problems. technical problem

Due to the different types of software engineering tasks, the developer portraits established by these technologies cannot provide recommendations for the comprehensive completion of software engineering tasks

Images