Graph-based code semantics learning for efficient semantic code clone detection

Yu, Dongjin; Yang, Quanxin; Chen, Xin; Chen, Jie; Xu, Yihang

doi:10.1016/j.infsof.2022.107130

Cited by 9 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The PDG was initially used in various code optimizations, such as parallelism detection [42], code movement, or program slicing [43]. This representation can also support program comprehension and maintenance by depicting statement dependency trees in order to assess change impact [44], find code similarities [45]- [48], measure code coverage, or reduce the test set for programs. Another graph-based code model that combines the Control Flow Graph (CFG) and Program Dependence Graph (PDG), along with an Abstract Syntax Tree, is known as the Code Property Graph (CPG) [49].…”

Section: A Code Dependency Representationsmentioning

confidence: 99%

Semantic Code Graph—An Information Model to Facilitate Software Comprehension

Borowski,

Balis,

Orzechowski

2024

IEEE Access

View full text Add to dashboard Cite

Software comprehension can be extremely time-consuming due to the ever-growing size of codebases. Consequently, there is an increasing need to accelerate the code comprehension process to facilitate maintenance and reduce associated costs. A crucial aspect of this process is understanding and preserving the high quality of the code dependency structure. While a variety of code structure models already exist, there is a surprising lack of models that closely represent the source code and focus on software comprehension. As a result, there are no readily available and easy-to-use tools to assist with dependency comprehension, refactoring, and quality monitoring of code. To address this gap, we propose the Semantic Code Graph (SCG), an information model that offers a detailed abstract representation of code dependencies with a close relationship to the source code. We establish the critical properties of the SCG model and demonstrate its implementation for Java and Scala languages. To validate the SCG model's usefulness in software comprehension, we compare it to nine other source code representation models. Additionally, we select 11 well-known and widelyused open-source projects developed in Java and Scala and perform a range of software comprehension activities on them using three different code representation models: the proposed SCG, the Call Graph, and the Class Collaboration Network. We then qualitatively analyze the results to compare the performance of these models in terms of software comprehension capabilities. These activities encompass project structure comprehension, identifying critical project entities, interactive visualization of code dependencies, and uncovering code similarities through software mining. Our findings demonstrate that the SCG enhances software comprehension capabilities compared to the prevailing Class Collaboration Network and Call Graph models. Moreover, the SCG-based data analysis yields actionable software comprehension insights. We also release an open-source tool, scg-cli, to assist with result reproduction and further research. We believe that the work described is a step towards the next generation of tools that streamline code dependency comprehension and management.

show abstract

Section: A Code Dependency Representationsmentioning

confidence: 99%

Semantic Code Graph—An Information Model to Facilitate Software Comprehension

Borowski,

Balis,

Orzechowski

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Type • CodeGraph4CCDetector (CG4CD). 36 CodeGraph4CCDetector use a Siamese graph-matching network based on attention mechanisms for CFGs/ PDGs, which can capture key information at the token, statement, and graph levels to sufficiently learn code semantics for code clone detection.…”

Section: Not Clone 19987mentioning

confidence: 99%

“…This method has achieved state‐of‐the‐art performance on code clone detection task. We use the same parameter settings released by the author to reproduce the results. CodeGraph4CCDetector (CG4CD) 36 . CodeGraph4CCDetector use a Siamese graph‐matching network based on attention mechanisms for CFGs/PDGs, which can capture key information at the token, statement, and graph levels to sufficiently learn code semantics for code clone detection. GRRLN .…”

Section: Experiments Setup and Research Questionsmentioning

confidence: 99%

GRRLN: Gated Recurrent Residual Learning Networks for code clone detection

Zhang,

Liu,

Shi

2024

J Software Evolu Process

View full text Add to dashboard Cite

Code clone detection is a critical problem in software development and maintenance domains. It aims to identify functionally identical or similar code fragments within an application. Existing works formulate the code clone detection task as a binary classification problem which predicts a code pair as a clone or not based on a pre‐defined threshold. In reality, there are various types of code clone subject to the degree of how a pair of code fragments are similar to each other. To investigate the effect of different code clone detection manners on the clone detection result, we propose Gated Recurrent Residual Learning Networks (GRRLN), a novel neural network model for code clone detection. To train GRRLN, we first represent each code fragment as a statement‐level tree sequence derived from the whole abstract syntax tree (AST). Then, a gated recurrent neural network with residual connections is adopted to fully extract the semantics of all individual statement trees together with their dependency relationships across the input statement sequence. Finally, the output representations of code fragments by GRRLN are used for similarity calculation and clone detection. We evaluate GRRLN using two real‐world datasets for code clone detection and clone type classification. Experiments show that GRRLN achieves promising and compelling results and meanwhile needs significantly less time and memory consumption compared with the state‐of‐the‐art methods.

show abstract

“…Theoretically, our choice of code2vec For source code representation is inspired by the recent performance analysis in method name prediction, code semantic extraction [34]. However, we cannot be sure whether the heavy computation and the cost of employing code2vec are worthy trade-offs.…”

Section: Why Does Attentivebuglocator Work?mentioning

confidence: 99%

AttentiveBugLocator: A Bug Localization Model using Attention-based SemanticFeatures and Information Retrieval

Ahmad,

Yu,

Kholief

et al. 2023

Preprint

View full text Add to dashboard Cite

In recent years, deep learning-based algorithms such as CNN, LSTM, and autoencoders have been proposed to rank suspicious buggy files. Meanwhile,representational learning has served to be the best approach to extract rich semantic features of bug reports and source code to reduce their lexical mismatch. In this paper, we propose AttentiveBugLocator, a Siamese-based representational learning modelfor improved bug localization performance. AttentiveBugLocator employs BERT and code2vec embedding models to produce richer semantic representations and a Siamese BiLSTM network with context attention to learn semantic matching between BRs and SFs. To further improve the effectiveness of AttentiveBugLocator, the semantic matching features are carefully fused with VSM, stack trace, and code complexity features. Evaluation results on four data sets show that AttentiveBugLocator can identify buggy files on the scale of 56% and 62% on MAP and MRR – thus, outperforms several state-of-the-art approaches.

show abstract

Graph-based code semantics learning for efficient semantic code clone detection

Cited by 9 publications

References 21 publications

Semantic Code Graph—An Information Model to Facilitate Software Comprehension

Semantic Code Graph—An Information Model to Facilitate Software Comprehension

GRRLN: Gated Recurrent Residual Learning Networks for code clone detection

AttentiveBugLocator: A Bug Localization Model using Attention-based SemanticFeatures and Information Retrieval

Contact Info

Product

Resources

About