Multi-Perspective Context Aggregation for Semi-supervised Cloze-style Reading Comprehension
This work addresses reading comprehension for natural language understanding, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles the problem of cloze-style reading comprehension by proposing a multi-perspective framework and an efficient sampling mechanism to address insufficient labeled data, achieving new state-of-the-art performance on the CLOTH dataset with nearly 100k questions.
Cloze-style reading comprehension has been a popular task for measuring the progress of natural language understanding in recent years. In this paper, we design a novel multi-perspective framework, which can be seen as the joint training of heterogeneous experts and aggregate context information from different perspectives. Each perspective is modeled by a simple aggregation module. The outputs of multiple aggregation modules are fed into a one-timestep pointer network to get the final answer. At the same time, to tackle the problem of insufficient labeled data, we propose an efficient sampling mechanism to automatically generate more training examples by matching the distribution of candidates between labeled and unlabeled data. We conduct our experiments on a recently released cloze-test dataset CLOTH (Xie et al., 2017), which consists of nearly 100k questions designed by professional teachers. Results show that our method achieves new state-of-the-art performance over previous strong baselines.