Gated-Attention Readers for Text Comprehension

Bhuwan Dhingra, Hanxiao Liu, Zhilin Yang, William W. Cohen, Ruslan Salakhutdinov
6/5/2016
cs.CLcs.LG

Official code: bdhingra/ga-reader (190 stars, BSD-2-Clause, last commit 2017-11-29).

Abstract

In this paper we study the problem of answering cloze-style questions over documents. Our model, the Gated-Attention (GA) Reader, integrates a multi-hop architecture with a novel attention mechanism, which is based on multiplicative interactions between the query embedding and the intermediate states of a recurrent neural network document reader. This enables the reader to build query-specific representations of tokens in the document for accurate answer selection. The GA Reader obtains state-of-the-art results on three benchmarks for this task--the CNN \& Daily Mail news stories and the Who Did What dataset. The effectiveness of multiplicative interaction is demonstrated by an ablation study, and by comparing to alternative compositional operators for implementing the gated-attention. The code is available at https://github.com/bdhingra/ga-reader.

View on arXivView PDF

Code Implementations(1)

bdhingra/ga-readerOfficial100%
BSD-2-Clause

Gated Attention Reader for Text Comprehension

19033PythonNov 4, 20168 years agoBSD-2-Clause

Similar Papers

Text Understanding with the Attention Sum Reader Network

82% match1 repo982016

Cognitive Graph for Multi-Hop Reading Comprehension at Scale

81% match1 repo4572019

Reading Wikipedia to Answer Open-Domain Questions

81% match1 repo4.5k2017

Graph Attention with Hierarchies for Multi-hop Question Answering

80% match2023

Constructing Cloze Questions Generatively

80% match2024

Query-Aware Graph Neural Networks for Enhanced Retrieval-Augmented Generation

80% match2025

Cite this paper

@article{dhingra2016gatedattention,
  title  = {Gated-Attention Readers for Text Comprehension},
  author = {Bhuwan Dhingra and Hanxiao Liu and Zhilin Yang and William W. Cohen and Ruslan Salakhutdinov},
  year   = {2016},
  eprint = {1606.01549},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1606.01549}
}

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/1606.01549.svg)](https://codeofpaper.com/paper/1606.01549)

Discussion