R$^3$: Reinforced Reader-Ranker for Open-Domain Question Answering

TL;DR

This paper introduces R$^3$, a novel open-domain question answering system that combines a passage ranker and a reader trained via reinforcement learning, significantly improving performance over previous methods.

Contribution

The paper presents a new pipeline with a learned passage ranker and a jointly trained reader using reinforcement learning for open-domain QA, advancing the state of the art.

Findings

Significant performance improvements on multiple open-domain QA datasets.

Effective joint training of ranker and reader with reinforcement learning.

Outperforms previous state-of-the-art methods.

Abstract

In recent years researchers have achieved considerable success applying neural network methods to question answering (QA). These approaches have achieved state of the art results in simplified closed-domain settings such as the SQuAD (Rajpurkar et al., 2016) dataset, which provides a pre-selected passage, from which the answer to a given question may be extracted. More recently, researchers have begun to tackle open-domain QA, in which the model is given a question and access to a large corpus (e.g., wikipedia) instead of a pre-selected passage (Chen et al., 2017a). This setting is more complex as it requires large-scale search for relevant passages by an information retrieval component, combined with a reading comprehension model that "reads" the passages to generate an answer to the question. Performance in this setting lags considerably behind closed-domain performance. In this paper, we present a novel open-domain QA system called Reinforced Ranker-Reader $(R^3)$, based on two algorithmic innovations. First, we propose a new pipeline for open-domain QA with a Ranker component, which learns to rank retrieved passages in terms of likelihood of generating the ground-truth answer to a given question. Second, we propose a novel method that jointly trains the Ranker along with an answer-generation Reader model, based on reinforcement learning. We report extensive experimental results showing that our method significantly improves on the state of the art for multiple open-domain QA datasets.