Message Passing Attention Networks for Document Understanding

Giannis Nikolentzos, Antoine J. -P. Tixier, Michalis Vazirgiannis
8/17/2019
cs.CL

Official code: giannisnik/mpad (61 stars, last commit 2021-03-11).

Abstract

Graph neural networks have recently emerged as a very effective framework for processing graph-structured data. These models have achieved state-of-the-art performance in many tasks. Most graph neural networks can be described in terms of message passing, vertex update, and readout functions. In this paper, we represent documents as word co-occurrence networks and propose an application of the message passing framework to NLP, the Message Passing Attention network for Document understanding (MPAD). We also propose several hierarchical variants of MPAD. Experiments conducted on 10 standard text classification datasets show that our architectures are competitive with the state-of-the-art. Ablation studies reveal further insights about the impact of the different components on performance. Code is publicly available at: https://github.com/giannisnik/mpad .

View on arXivView PDF

Code Implementations(1)

giannisnik/mpadOfficial100%

Message Passing Attention Networks for Document Understanding

6117PythonJul 24, 20195 years ago
attention-mechanismgraph-neural-networkshierarchical-modelstext-classificationtext-mining

Similar Papers

Graph Neural Networks for Text Classification: A Survey

85% match2023

Tensor Graph Convolutional Networks for Text Classification

85% match1 repo292020

Connecting the Dots: What Graph-Based Text Representations Work Best for Text Classification Using Graph Neural Networks?

85% match1 repo22023

Graph Convolutional Networks for Text Classification

84% match1 repo1.4k2018

Graph Coordinates and Conventional Neural Networks -- An Alternative for Graph Neural Networks

84% match2023

Text Level Graph Neural Network for Text Classification

84% match1 repo522019

Cite this paper

@article{nikolentzos2019message,
  title  = {Message Passing Attention Networks for Document Understanding},
  author = {Giannis Nikolentzos and Antoine J. -P. Tixier and Michalis Vazirgiannis},
  year   = {2019},
  eprint = {1908.06267},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1908.06267}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1908.06267.svg)](https://codeofpaper.com/paper/1908.06267)

Discussion