Inferring Javascript types using Graph Neural Networks
arXiv:1905.06707v17 citations
Originality Synthesis-oriented
AI Analysis
This work addresses automatic code repair for programmers, but it is incremental as it builds on existing methods for type inference.
The paper tackled the problem of predicting token types in Javascript programs using a graph neural network model, achieving an accuracy above 90% and improving on previous similar work.
The recent use of `Big Code' with state-of-the-art deep learning methods offers promising avenues to ease program source code writing and correction. As a first step towards automatic code repair, we implemented a graph neural network model that predicts token types for Javascript programs. The predictions achieve an accuracy above $90\%$, which improves on previous similar work.