Graph Neural Networks for Natural Language Processing

The repository contains code examples for GNN-for-NLP tutorial at EMNLP 2019 and CODS-COMAD 2020.

Slides can be downloaded from here.

Dependencies

• Compatible with PyTorch 1.x, TensorFlow 1.x and Python 3.x.
• Dependencies can be installed using requirements.txt.

TensorFlow Examples:

• tf_gcn.py contains simplified implementation of first-order approximation of GCN model proposed by Kipf et. al. (2016)
• Extensions of the same implementation for different problems:
  • Relation Extraction: RESIDE
  • GCNs for Word Embeddings: WordGCN
  • Document Time-stamping: NeuralDater

PyTorch Examples:

• pytorch_gcn.py is pytorch equivalent of tf_gcn.py implemented using pytorch-geometric.
• Several other examples are available here.

Additional Resources:

• Short writeup on theory behind Graph Convolutional Networks [Pdf] (refer Chapter-2).
• GNN recent papers.

Citation:

@inproceedings{vashishth-etal-2019-graph,
    title = "Graph-based Deep Learning in Natural Language Processing",
    author = "Vashishth, Shikhar  and
      Yadati, Naganand  and
      Talukdar, Partha",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP): Tutorial Abstracts",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    abstract = "This tutorial aims to introduce recent advances in graph-based deep learning techniques such as Graph Convolutional Networks (GCNs) for Natural Language Processing (NLP). It provides a brief introduction to deep learning methods on non-Euclidean domains such as graphs and justifies their relevance in NLP. It then covers recent advances in applying graph-based deep learning methods for various NLP tasks, such as semantic role labeling, machine translation, relationship extraction, and many more.",
}