Multilingual training for Software Engineering

TL;DR

This paper demonstrates that multilingual training data, leveraging similarities across programming languages, can enhance machine learning models for software engineering tasks like code summarization, retrieval, and naming.

Contribution

It introduces a novel approach of using cross-language similarities, especially in identifiers, to augment training data and improve model performance in software engineering tasks.

Findings

Multilingual code data shows high similarity, especially in identifiers.

Using multilingual data improves performance across tasks.

Identifier patterns are crucial for training data effectiveness.

Abstract

Well-trained machine-learning models, which leverage large amounts of open-source software data, have now become an interesting approach to automating many software engineering tasks. Several SE tasks have all been subject to this approach, with performance gradually improving over the past several years with better models and training methods. More, and more diverse, clean, labeled data is better for training; but constructing good-quality datasets is time-consuming and challenging. Ways of augmenting the volume and diversity of clean, labeled data generally have wide applicability. For some languages (e.g., Ruby) labeled data is less abundant; in others (e.g., JavaScript) the available data maybe more focused on some application domains, and thus less diverse. As a way around such data bottlenecks, we present evidence suggesting that human-written code in different languages (which performs the same function), is rather similar, and particularly preserving of identifier naming patterns; we further present evidence suggesting that identifiers are a very important element of training data for software engineering tasks. We leverage this rather fortuitous phenomenon to find evidence that available multilingual training data (across different languages) can be used to amplify performance. We study this for 3 different tasks: code summarization, code retrieval, and function naming. We note that this data-augmenting approach is broadly compatible with different tasks, languages, and machine-learning models.