Building Dynamic Knowledge Graphs from Text-based Games

Abstract

We are interested in learning how to update Knowledge Graphs (KG) from text. In this preliminary work, we propose a novel Sequence-to-Sequence (Seq2Seq) architecture to generate elementary KG operations. Furthermore, we introduce a new dataset for KG extraction built upon text-based game transitions (over 300k data points). We conduct experiments and discuss the results.

NeurIPS 2019, Graph Representation Learning (GRL) Workshop

References in corpus (6)

• TextWorld: A Learning Environment for Text-based Games
• Pointing the Unknown Words
• Building Dynamic Knowledge Graphs from Text using Machine Reading Comprehension
• One-Shot Relational Learning for Knowledge Graphs
• Does Order Matter? An Empirical Study on Generating Multiple Keyphrases as a Sequence
• Playing Text-Adventure Games with Graph-Based Deep Reinforcement Learning

Cited by in corpus (3)

• Learning Dynamic Belief Graphs to Generalize on Text-Based Games
• Deep Reinforcement Learning with Stacked Hierarchical Attention for Text-based Games
• WordCraft: An Environment for Benchmarking Commonsense Agents