Learning Open Domain Multi-hop Search Using Reinforcement Learning

TL;DR

This paper introduces a reinforcement learning approach for training an agent to efficiently perform multi-hop relation searches in open domains, reducing document processing while maintaining accuracy.

Contribution

It formulates the multi-hop search as a Markov decision process and employs actor-critic reinforcement learning to learn effective search policies.

Findings

The method reduces the number of documents processed compared to baselines.

It successfully extracts multi-hop paths in Wikipedia-based search problems.

The learned policies outperform heuristic algorithms in efficiency.

Abstract

We propose a method to teach an automated agent to learn how to search for multi-hop paths of relations between entities in an open domain. The method learns a policy for directing existing information retrieval and machine reading resources to focus on relevant regions of a corpus. The approach formulates the learning problem as a Markov decision process with a state representation that encodes the dynamics of the search process and a reward structure that minimizes the number of documents that must be processed while still finding multi-hop paths. We implement the method in an actor-critic reinforcement learning algorithm and evaluate it on a dataset of search problems derived from a subset of English Wikipedia. The algorithm finds a family of policies that succeeds in extracting the desired information while processing fewer documents compared to several baseline heuristic algorithms.