# Open Vocabulary Learning on Source Code with a Graph-Structured Cache

arXiv: https://arxiv.org/abs/1810.08305 · HTML: https://mlsanity.com/paper/1810.08305/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Oct 2018.

## Built on by (most cited first, 10 of 10)

- [Code Structure Guided Transformer for Source Code Summarization](https://mlsanity.com/paper/2104.09340/) (2021)
- [A Survey on Open Set Recognition](https://mlsanity.com/paper/2109.00893/) (2021)
- [A Survey on Machine Learning Techniques for Source Code Analysis](https://mlsanity.com/paper/2110.09610/) (2021)
- [Deep Learning for Code Intelligence: Survey, Benchmark and Toolkit](https://mlsanity.com/paper/2401.00288/) (2023)
- [GraphSearchNet: Enhancing GNNs via Capturing Global Dependencies for Semantic Code Search](https://mlsanity.com/paper/2111.02671/) (2021)
- [GypSum: Learning Hybrid Representations for Code Summarization](https://mlsanity.com/paper/2204.12916/) (2022)
- [A Neural Network Architecture for Program Understanding Inspired by Human Behaviors](https://mlsanity.com/paper/2206.04730/) (2022)
- [A Survey of Deep Learning Models for Structural Code Understanding](https://mlsanity.com/paper/2205.01293/) (2022)
- [Learning to Extend Program Graphs to Work-in-Progress Code](https://mlsanity.com/paper/2105.14038/) (2021)
- [Precise Learning of Source Code Contextual Semantics via Hierarchical Dependence Structure and Graph Attention Networks](https://mlsanity.com/paper/2111.11435/) (2021)

## Abstract (arXiv)

Machine learning models that take computer program source code as input typically use Natural Language Processing (NLP) techniques. However, a major challenge is that code is written using an open, rapidly changing vocabulary due to, e.g., the coinage of new variable and method names. Reasoning over such a vocabulary is not something for which most NLP methods are designed. We introduce a Graph-Structured Cache to address this problem; this cache contains a node for each new word the model encounters with edges connecting each word to its occurrences in the code. We find that combining this graph-structured cache strategy with recent Graph-Neural-Network-based models for supervised learning on code improves the models' performance on a code completion task and a variable naming task --- with over 100% relative improvement on the latter --- at the cost of a moderate increase in computation time.