Deep code comment generation

TLDR

Code comments aid comprehension but are frequently mismatched, missing, or outdated, forcing developers to infer functionality from source code. The authors propose DeepCom, an automatic code‑comment generation system for Java methods, to aid developers in understanding method functionality. DeepCom learns from a large corpus of Java code using NLP and a deep neural network that analyzes method structure, and its performance is evaluated with a machine‑translation metric on data from 9,714 GitHub projects. Experimental results show DeepCom outperforms state‑of‑the‑art methods by a substantial margin.

Abstract

During software maintenance, code comments help developers comprehend programs and reduce additional time spent on reading and navigating source code. Unfortunately, these comments are often mismatched, missing or outdated in the software projects. Developers have to infer the functionality from the source code. This paper proposes a new approach named DeepCom to automatically generate code comments for Java methods. The generated comments aim to help developers understand the functionality of Java methods. DeepCom applies Natural Language Processing (NLP) techniques to learn from a large code corpus and generates comments from learned features. We use a deep neural network that analyzes structural information of Java methods for better comments generation. We conduct experiments on a large-scale Java corpus built from 9,714 open source projects from GitHub. We evaluate the experimental results on a machine translation metric. Experimental results demonstrate that our method DeepCom outperforms the state-of-the-art by a substantial margin.