VarCLR: Variable Semantic Representation Pre-training via Contrastive Learning

Download: Paper, ACM Link, Video.

“VarCLR: Variable Semantic Representation Pre-training via Contrastive Learning” by Qibin Chen, Jeremy Lacomis, Edward J. Schwartz, Graham Neubig, Bogdan Vasilescu, and Claire Le Goues. In Proceedings of the International Conference on Software Engineering, 2022.

Abstract

Variable names are critical for conveying intended program behavior. Machine learning-based program analysis methods use variable name representations for a wide range of tasks, such as suggesting new variable names and bug detection. Ideally, such methods could capture semantic relationships between names beyond syntactic similarity, e.g., the fact that the names average and mean are similar. Unfortunately, previous work has found that even the best of previous representation approaches primarily capture “relatedness” (whether two variables are linked at all), rather than “similarity” (whether they actually have the same meaning).

We propose VarCLR, a new approach for learning semantic representations of variable names that effectively captures variable similarity in this stricter sense. We observe that this problem is an excellent fit for contrastive learning, which aims to minimize the distance between explicitly similar inputs, while maximizing the distance between dissimilar inputs. This requires labeled training data, and thus we construct a novel, weakly-supervised variable renaming dataset mined from GitHub edits. We show that VarCLR enables the effective application of sophisticated, general-purpose language models like BERT, to variable name representation and thus also to related downstream tasks like variable name similarity search or spelling correction. VarCLR produces models that significantly outperform the state-of-the-art on IdBench, an existing benchmark that explicitly captures variable similarity (as distinct from relatedness). Finally, we contribute a release of all data, code, and pre-trained models, aiming to provide a plug-in replacement for variable representations used in either existing or future program analyses that rely on variable names.

Download: Paper, ACM Link, Video.

BibTeX entry:

@inproceedings{chen:2022:icse,
   author = {Qibin Chen and Jeremy Lacomis and Edward J. Schwartz and
	Graham Neubig and Bogdan Vasilescu and Claire Le Goues},
   title = {VarCLR: Variable Semantic Representation Pre-training via
	Contrastive Learning},
   booktitle = {Proceedings of the International Conference on Software
	Engineering},
   year = {2022}
}

(This webpage was created with bibtex2web.)

Back to publications.