Compositionality-Aware Graph2Seq Learning
This work addresses the challenge of improving model efficiency and performance in graph-to-sequence tasks, particularly for source code analysis, but it is incremental as it builds on existing GNN methods.
The paper tackles the problem of generating human-interpretable sequences from graphs, specifically in extreme source code summarization, and demonstrates that a compositionality-aware GNN architecture outperforms the previous state-of-the-art model with over seven times fewer parameters.
Graphs are a highly expressive data structure, but it is often difficult for humans to find patterns from a complex graph. Hence, generating human-interpretable sequences from graphs have gained interest, called graph2seq learning. It is expected that the compositionality in a graph can be associated to the compositionality in the output sequence in many graph2seq tasks. Therefore, applying compositionality-aware GNN architecture would improve the model performance. In this study, we adopt the multi-level attention pooling (MLAP) architecture, that can aggregate graph representations from multiple levels of information localities. As a real-world example, we take up the extreme source code summarization task, where a model estimate the name of a program function from its source code. We demonstrate that the model having the MLAP architecture outperform the previous state-of-the-art model with more than seven times fewer parameters than it.