“Assembling Information from Big Corpora by Focusing Machine Reading”, presented by Enrique Noriega
We propose a framework to teach an automated agent to learn how to search for multi-hop paths of relations between entities in large corpora. 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 framework with reinforcement learning and evaluate it on an open-domain dataset of search problems derived from a subset of English Wikipedia using a policy gradient actor-critic algorithm and a domain-specific dataset of search problems in the biomedical domain using a temporal-difference learning algorithm.
We show that deploying the focused reading framework with reinforcement learning finds policies that retrieve more multi-hop paths while processing fewer documents compared to several strong deterministic baseline implementations.
