Constructing Multilingual Code Search Dataset Using Neural Machine Translation

TL;DR

This paper introduces a multilingual code search dataset created via neural machine translation, enabling better training of models across multiple natural and programming languages, and evaluates the impact of data quality and size.

Contribution

It presents a new multilingual dataset for code search in four natural and four programming languages using neural machine translation.

Findings

Pre-trained models perform best with combined multilingual data.

Back-translation filtering influences model performance.

Data size impacts model effectiveness more than translation quality.

Abstract

Code search is a task to find programming codes that semantically match the given natural language queries. Even though some of the existing datasets for this task are multilingual on the programming language side, their query data are only in English. In this research, we create a multilingual code search dataset in four natural and four programming languages using a neural machine translation model. Using our dataset, we pre-train and fine-tune the Transformer-based models and then evaluate them on multiple code search test sets. Our results show that the model pre-trained with all natural and programming language data has performed best in most cases. By applying back-translation data filtering to our dataset, we demonstrate that the translation quality affects the model's performance to a certain extent, but the data size matters more.