A Convolutional Attention Network for Extreme Summarization of Source Code

TL;DR

This paper introduces a convolutional attention neural network designed for extreme summarization of source code, effectively capturing local and long-range features to generate concise, descriptive summaries of code snippets.

Contribution

The novel convolutional attention architecture improves code summarization by combining local feature detection with attention mechanisms, outperforming previous models on Java projects.

Findings

Achieves better performance than previous attention models.

Effectively captures local and long-range features in code.

Generates concise, accurate code summaries.

Abstract

Attention mechanisms in neural networks have proved useful for problems in which the input and output do not have fixed dimension. Often there exist features that are locally translation invariant and would be valuable for directing the model's attention, but previous attentional architectures are not constructed to learn such features specifically. We introduce an attentional neural network that employs convolution on the input tokens to detect local time-invariant and long-range topical attention features in a context-dependent way. We apply this architecture to the problem of extreme summarization of source code snippets into short, descriptive function name-like summaries. Using those features, the model sequentially generates a summary by marginalizing over two attention mechanisms: one that predicts the next summary token based on the attention weights of the input tokens and another that is able to copy a code token as-is directly into the summary. We demonstrate our convolutional attention neural network's performance on 10 popular Java projects showing that it achieves better performance compared to previous attentional mechanisms.