Using a Nearest-Neighbour, BERT-Based Approach for Scalable Clone Detection

TL;DR

This paper introduces SSCD, a scalable BERT-based clone detection method that efficiently identifies inexact code clones in large codebases by using embeddings and nearest neighbor search, outperforming existing approaches.

Contribution

The paper presents SSCD, a novel BERT-based clone detection approach that improves scalability and effectiveness for detecting complex code clones in industrial-scale datasets.

Findings

SSCD outperforms state-of-the-art clone detection tools like SAGA and SourcererCC.

SSCD can process 320 million lines of code in under three hours.

Shorter input lengths and text-only models enhance efficiency with minimal effectiveness loss.

Abstract

Code clones can detrimentally impact software maintenance and manually detecting them in very large codebases is impractical. Additionally, automated approaches find detection of Type 3 and Type 4 (inexact) clones very challenging. While the most recent artificial deep neural networks (for example BERT-based artificial neural networks) seem to be highly effective in detecting such clones, their pairwise comparison of every code pair in the target system(s) is inefficient and scales poorly on large codebases. We therefore introduce SSCD, a BERT-based clone detection approach that targets high recall of Type 3 and Type 4 clones at scale (in line with our industrial partner's requirements). It does so by computing a representative embedding for each code fragment and finding similar fragments using a nearest neighbour search. SSCD thus avoids the pairwise-comparison bottleneck of other Neural Network approaches while also using parallel, GPU-accelerated search to tackle scalability. This paper details the approach and an empirical assessment towards configuring and evaluating that approach in industrial setting. The configuration analysis suggests that shorter input lengths and text-only based neural network models demonstrate better efficiency in SSCD, while only slightly decreasing effectiveness. The evaluation results suggest that SSCD is more effective than state-of-the-art approaches like SAGA and SourcererCC. It is also highly efficient: in its optimal setting, SSCD effectively locates clones in the entire 320 million LOC BigCloneBench (a standard clone detection benchmark) in just under three hours.