On the Use of Deep Learning Models for Semantic Clone Detection

TL;DR

This paper evaluates deep learning models for semantic clone detection across multiple datasets, revealing insights into their robustness and generalizability, especially highlighting the strengths of cross-language models like C4.

Contribution

It introduces a multi-step evaluation framework for clone detection models across diverse datasets, including a new GPT-assisted dataset, and compares their robustness using mutation operators.

Findings

Single-language models perform well on BigCloneBench but vary on SemanticCloneBench.

Cross-language model C4 outperforms others on SemanticCloneBench and shows consistent robustness.

Mutation testing reveals C4's superior stability across datasets.

Abstract

Detecting and tracking code clones can ease various software development and maintenance tasks when changes in a code fragment should be propagated over all its copies. Several deep learning-based clone detection models have appeared in the literature for detecting syntactic and semantic clones, widely evaluated with the BigCloneBench dataset. However, class imbalance and the small number of semantic clones make BigCloneBench less ideal for interpreting model performance. Researchers also use other datasets such as GoogleCodeJam, OJClone, and SemanticCloneBench to understand model generalizability. To overcome the limitations of existing datasets, the GPT-assisted semantic and cross-language clone dataset GPTCloneBench has been released. However, how these models compare across datasets remains unclear. In this paper, we propose a multi-step evaluation approach for five state-of-the-art clone detection models leveraging existing benchmark datasets, including GPTCloneBench, and using mutation operators to study model ability. Specifically, we examine three highly-performing single-language models (ASTNN, GMN, CodeBERT) on BigCloneBench, SemanticCloneBench, and GPTCloneBench, testing their robustness with mutation operations. Additionally, we compare them against cross-language models (C4, CLCDSA) known for detecting semantic clones. While single-language models show high F1 scores for BigCloneBench, their performance on SemanticCloneBench varies (up to 20%). Interestingly, the cross-language model (C4) shows superior performance (around 7%) on SemanticCloneBench over other models and performs similarly on BigCloneBench and GPTCloneBench. On mutation-based datasets, C4 has more robust performance (less than 1% difference) compared to single-language models, which show high variability.