Teaching Machine Comprehension with Compositional Explanations
arXiv:2005.00806
Abstract
Advances in machine reading comprehension (MRC) rely heavily on the collection of large scale human-annotated examples in the form of (question, paragraph, answer) triples. In contrast, humans are typically able to generalize with only a few examples, relying on deeper underlying world knowledge, linguistic sophistication, and/or simply superior deductive powers. In this paper, we focus on "teaching" machines reading comprehension, using a small number of semi-structured explanations that explicitly inform machines why answer spans are correct. We extract structured variables and rules from explanations and compose neural module teachers that annotate instances for training downstream MRC models. We use learnable neural modules and soft logic to handle linguistic variation and overcome sparse coverage; the modules are jointly optimized with the MRC model to improve final performance. On the SQuAD dataset, our proposed method achieves 70.14% F1 score with supervision from 26 explanations, comparable to plain supervised learning using 1,100 labeled instances, yielding a 12x speed up.
arXiv commentAccepted to EMNLP 2020 Findings. Camera-ready version. Project page: http://inklab.usc.edu/mrc-explanation-project/
References in corpus (8)
- Unsupervised Data Augmentation for Consistency Training
- Bidirectional Attention Flow for Machine Comprehension
- Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
- Learning to Map Sentences to Logical Form: Structured Classification with Probabilistic Categorial Grammars
- Unsupervised Question Answering by Cloze Translation
- KERMIT: Generative Insertion-Based Modeling for Sequences
- Neural Module Networks for Reasoning over Text
- Learning from Explanations with Neural Execution Tree