Enriching query semantics for code search with reinforcement learning. (January 2022)

• Record Type: Journal Article
• Authors: Wang, Chaozheng; Nong, Zhenhao; Gao, Cuiyun; Li, Zongjie; Zeng, Jichuan; Xing, Zhenchang; Liu, Yang
• Is Part Of: Neural networks. Volume 145(2022)
• Page Start: 22
• Page End: 32
• Publication Date: 2022-01
• Subjects: Code search -- Query semantics -- Semantic enrichment -- Reinforcement learning; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2021.09.025
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: Code search is a common practice for developers during software implementation. The challenges of accurate code search mainly lie in the knowledge gap between source code and natural language (i.e., queries). Due to the limited code-query pairs and large code-description pairs available, the prior studies based on deep learning techniques focus on learning the semantic matching relation between source code and corresponding description texts for the task, and hypothesize that the semantic gap between descriptions and user queries is marginal. In this work, we found that the code search models trained on code-description pairs may not perform well on user queries, which indicates the semantic distance between queries and code descriptions. To mitigate the semantic distance for more effective code search, we propose QueCos, a Que ry-enriched Co de s earch model. QueCos learns to generate semantic enriched queries to capture the key semantics of given queries with reinforcement learning (RL). With RL, the code search performance is considered as a reward for producing accurate semantic enriched queries. The enriched queries are finally employed for code search. Experiments on the benchmark datasets show that QueCos can significantly outperform the state-of-the-art code search models.