Towards the Use of Deep Reinforcement Learning with Global Policy For Query-based Extractive Summarisation
This is an incremental improvement for text summarization researchers, addressing a specific bottleneck in supervised methods.
The paper tackles the mismatch between sentence-level labels and summary-level evaluation in query-based extractive summarization by using deep reinforcement learning with a global policy, achieving encouraging results.
Supervised approaches for text summarisation suffer from the problem of mismatch between the target labels/scores of individual sentences and the evaluation score of the final summary. Reinforcement learning can solve this problem by providing a learning mechanism that uses the score of the final summary as a guide to determine the decisions made at the time of selection of each sentence. In this paper we present a proof-of-concept approach that applies a policy-gradient algorithm to learn a stochastic policy using an undiscounted reward. The method has been applied to a policy consisting of a simple neural network and simple features. The resulting deep reinforcement learning system is able to learn a global policy and obtain encouraging results.