Exploiting Explicit Paths for Multi-hop Reading Comprehension

  title={Exploiting Explicit Paths for Multi-hop Reading Comprehension},
  author={Souvik Kundu and Tushar Khot and A. Sabharwal},
  • Souvik Kundu, Tushar Khot, A. Sabharwal
  • Published in ACL 2019
  • Computer Science
  • We propose a novel, path-based reasoning approach for the multi-hop reading comprehension task where a system needs to combine facts from multiple passages to answer a question. Although inspired by multi-hop reasoning over knowledge graphs, our proposed approach operates directly over unstructured text. It generates potential paths through passages and scores them without any direct path supervision. The proposed model, named PathNet, attempts to extract implicit relations from text through… CONTINUE READING
    16 Citations

    Figures, Tables, and Topics from this paper.

    Explore Further: Topics Discussed in This Paper

    A Sentence-Based Circular Reasoning Model in Multi-Hop Reading Comprehension
    Multi-hop Question Answering via Reasoning Chains
    • 17
    • PDF
    Multi-hop Reading Comprehension across Multiple Documents by Reasoning over Heterogeneous Graphs
    • 30
    • Highly Influenced
    • PDF
    Learning to Generate Multi-Hop Knowledge Paths for Commonsense Question Answering
    • 2020
    Scalable Multi-Hop Relational Reasoning for Knowledge-Aware Question Answering
    • 6
    • PDF
    Connecting the Dots: A Knowledgeable Path Generator for Commonsense Question Answering
    • 2
    • PDF


    Constructing Datasets for Multi-hop Reading Comprehension Across Documents
    • 209
    • Highly Influential
    • PDF
    Exploring Graph-structured Passage Representation for Multi-hop Reading Comprehension with Graph Neural Networks
    • 51
    • Highly Influential
    • PDF
    Question Answering by Reasoning Across Documents with Graph Convolutional Networks
    • 84
    • Highly Influential
    • PDF
    Question Answering as Global Reasoning Over Semantic Abstractions
    • 46
    • PDF
    Bidirectional Attention Flow for Machine Comprehension
    • 1,191
    • PDF
    Higher-order Lexical Semantic Models for Non-factoid Answer Reranking
    • 36
    • PDF
    Chains of Reasoning over Entities, Relations, and Text using Recurrent Neural Networks
    • 141
    • Highly Influential
    • PDF
    Traversing Knowledge Graphs in Vector Space
    • 234
    • PDF
    Semantic Parsing on Freebase from Question-Answer Pairs
    • 982
    • PDF