Learning Dynamic Knowledge Graphs to Generalize on Text-Based Games

02/21/2020
by   Ashutosh Adhikari, et al.
0

Playing text-based games requires skill in processing natural language and in planning. Although a key goal for agents solving this task is to generalize across multiple games, most previous work has either focused on solving a single game or has tackled generalization with rule-based heuristics. In this work, we investigate how structured information in the form of a knowledge graph (KG) can facilitate effective planning and generalization. We introduce a novel transformer-based sequence-to-sequence model that constructs a "belief" KG from raw text observations of the environment, dynamically updating this belief graph at every game step as it receives new observations. To train this model to build useful graph representations, we introduce and analyze a set of graph-related pre-training tasks. We demonstrate empirically that KG-based representations from our model help agents to converge faster to better policies for multiple text-based games, and further, enable stronger zero-shot performance on unseen games. Experiments on unseen games show that our best agent outperforms text-based baselines by 21.6

READ FULL TEXT

page 17

page 18

page 24

page 25

research
01/24/2020

Exploration Based Language Learning for Text-Based Games

This work presents an exploration and imitation-learning-based agent cap...
research
05/17/2018

Language Expansion In Text-Based Games

Text-based games are suitable test-beds for designing agents that can le...
research
02/16/2023

Learning Multi-Object Positional Relationships via Emergent Communication

The study of emergent communication has been dedicated to interactive ar...
research
10/21/2019

Building Dynamic Knowledge Graphs from Text-based Games

We are interested in learning how to update Knowledge Graphs (KG) from t...
research
12/03/2018

Towards Solving Text-based Games by Producing Adaptive Action Spaces

To solve a text-based game, an agent needs to formulate valid text comma...
research
06/29/2018

Counting to Explore and Generalize in Text-based Games

We propose a recurrent RL agent with an episodic exploration mechanism t...
research
09/24/2020

Bootstrapped Q-learning with Context Relevant Observation Pruning to Generalize in Text-based Games

We show that Reinforcement Learning (RL) methods for solving Text-Based ...

Please sign up or login with your details

Forgot password? Click here to reset