IECS: Intent-Enforced Code Search via Extended Boolean Model

Yang, Yangrui; Huang, Qing

doi:10.3233/jifs-161994

Cited by 10 publications

(5 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are mainly three stages in the development of code search models. Traditional information retrieval techniques match keywords between queries and code fragments (Hill et al, 2011;Yang and Huang, 2017;Satter and Sakib, 2016;Lv et al, 2015;Van Nguyen et al, 2017). Since natural language and programming language have different syntax rules, they often suffer from vocabulary mismatch problems (McMillan et al, 2011).…”

Section: Related Workmentioning

confidence: 99%

Rethinking Negative Pairs in Code Search

Li,

Zhou,

Luu

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Recently, contrastive learning has become a key component in fine-tuning code search models for software development efficiency and effectiveness. It pulls together positive code snippets while pushing negative samples away given search queries. Among contrastive learning, InfoNCE is the most widely used loss function due to its better performance. However, the following problems in negative samples of In-foNCE may deteriorate its representation learning: 1) The existence of false negative samples in large code corpora due to duplications. 2). The failure to explicitly differentiate between the potential relevance of negative samples. As an example, a bubble sorting algorithm example is less "negative" than a file saving function for the quick sorting algorithm query. In this paper, we tackle the above problems by proposing a simple yet effective Soft-InfoNCE loss that inserts weight terms into InfoNCE. In our proposed loss function, we apply three methods to estimate the weights of negative pairs and show that the vanilla InfoNCE loss is a special case of Soft-InfoNCE. Theoretically, we analyze the effects of Soft-InfoNCE on controlling the distribution of learnt code representations and on deducing a more precise mutual information estimation. We furthermore discuss the superiority of proposed loss functions with other design alternatives.Extensive experiments demonstrate the effectiveness of Soft-InfoNCE and weights estimation methods under state-of-the-art code search models on a large-scale public dataset consisting of six programming languages. Source code is available at https://github.com/ Alex-HaochenLi/Soft-InfoNCE.

show abstract

Section: Related Workmentioning

confidence: 99%

Rethinking Negative Pairs in Code Search

Li,

Zhou,

Luu

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

show abstract

“…However, the above models only connect query words with the Boolean operator "OR" implicitly and cannot address the "AND" operation. To support code search with complete Boolean queries, two studies [83,140] in Table 4 extended query with related words and the operator "AND", and leveraged the Extended Boolean Model (EBM) [113] to calculate the relevancy between query and code. We can notice that all the above models match the words in query and code directly.…”

Section: Modelsmentioning

confidence: 99%

“…It has been observed that APIs are an important factor to complement the missing semantics in queries [8]. Researchers have thus expanded query words with relevant APIs or class names from official API documents [83], codebases [140], or Stack Overflow posts [11,42,95,118,143].…”

Section: Classification Of Code Search Tasksmentioning

confidence: 99%

“…As illustrated in Table 17, 39% of queries were collected from question and answer forums including Stack Overflow [19] and Eclipse FAQs [8]; 6% of queries were extracted from software development kits, including JDK [38], Android development kit [144], and .Net framework [130]. The remaining 55% of queries were manually selected from frequently used examples: codebase [66,124], search logs of practically used code search engine [85,140], a summary of search on the Internet [90,104], and automatically generated queries [47,150].…”

Section: Queriesmentioning

confidence: 99%

“…The codebase scale in a practical environment (e.g., GitHub and FDroid) is usually large with millions of lines of code. However, many code search studies only tested their tools on a small scale codebase [43,70,130,140]. In this case, findings may not generalize to large codebases [70].…”

Section: Challenges and Opportunities 71 Challengesmentioning

confidence: 99%

See 2 more Smart Citations

Opportunities and Challenges in Code Search Tools

Liu,

Xia,

et al. 2020

Preprint

View full text Add to dashboard Cite

Code search is a core software engineering task. Effective code search tools can help developers substantially improve their software development efficiency and effectiveness. In recent years, many code search studies have leveraged different techniques, such as deep learning and information retrieval approaches, to retrieve expected code from a large-scale codebase. However, there is a lack of a comprehensive comparative summary of existing code search approaches. To understand the research trends in existing code search studies, we systematically reviewed 81 relevant studies. We investigated the publication trends of code search studies, analyzed key components, such as codebase, query, and modeling technique used to build code search tools, and classified existing tools into focusing on supporting seven different search tasks. Based on our findings, we identified a set of outstanding challenges in existing studies and a research roadmap for future code search research.CCS Concepts: • Software and its engineering → Search-based software engineering.

show abstract