A Novel SMOTE Algorithm based Portrait Model for Programmers
| Autor: | Pengyu Yu, Yirui Wu, Shun Zhao |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
User profile
business.industry Computer science Process (engineering) 020208 electrical & electronic engineering Information technology 02 engineering and technology Software quality Data modeling Software Encoding (memory) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing business Algorithm Coding (social sciences) |
| Zdroj: | HPCC/DSS/SmartCity |
| DOI: | 10.1109/hpcc-smartcity-dss50907.2020.00159 |
| Popis: | With the rapid development of information technology, the scale of software products is getting larger. In order to ensure code quality, it’s necessary to collect log information in the continuous integration development process, analyze the behavior of developers, build developer user profile models, and provide personalized suggestions to developers based on profile information. All these processes can be defined as goals of user portrait model. Essentially, user portrait model is designed to model users’ coding behaviours based on a large amount of data. However, current methods often suffer from the problem of imbalanced data, which is the core challenge for a successful portrait model. In fact, most parts of log information refer to regular programmers, while only few samples correspond to programmers who are supposed to be improved by suggestions from the portrait model. To solve this problem, we propose to adopt SMOTE Algorithm to deal with the imbalanced log data, which is the core innovation of the proposed method. Experiments show the proposed SMOTE Algorithm based model could accurately classify programmers’ personality types and offer suggestions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|