Learning to Update Natural Language Comments Based on Code Changes

Autor:	Panthaplackel, Sheena, Nie, Pengyu, Gligoric, Milos, Li, Junyi Jessy, Mooney, Raymond J.
Rok vydání:	2020
Předmět:	Computer Science - Computation and Language Computer Science - Machine Learning Computer Science - Software Engineering
Druh dokumentu:	Working Paper
Popis:	We formulate the novel task of automatically updating an existing natural language comment based on changes in the body of code it accompanies. We propose an approach that learns to correlate changes across two distinct language representations, to generate a sequence of edits that are applied to the existing comment to reflect the source code modifications. We train and evaluate our model using a dataset that we collected from commit histories of open-source software projects, with each example consisting of a concurrent update to a method and its corresponding comment. We compare our approach against multiple baselines using both automatic metrics and human evaluation. Results reflect the challenge of this task and that our model outperforms baselines with respect to making edits. Comment: Accepted in Association for Computational Linguistics (ACL) 2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2004.12169 Zobrazit plný text záznamu View this record from Arxiv