Towards Interpretable Natural Language Understanding with Explanations as Latent Variables

TL;DR

This paper introduces a framework for interpretable natural language understanding that uses explanations as latent variables, enabling effective predictions with limited annotated explanations and semi-supervised learning.

Contribution

It proposes a variational EM-based approach treating explanations as latent variables, reducing the need for extensive annotated data and enhancing interpretability and performance.

Findings

Effective in supervised and semi-supervised settings

Generates high-quality natural language explanations

Improves prediction accuracy with limited annotations

Abstract

Recently generating natural language explanations has shown very promising results in not only offering interpretable explanations but also providing additional information and supervision for prediction. However, existing approaches usually require a large set of human annotated explanations for training while collecting a large set of explanations is not only time consuming but also expensive. In this paper, we develop a general framework for interpretable natural language understanding that requires only a small set of human annotated explanations for training. Our framework treats natural language explanations as latent variables that model the underlying reasoning process of a neural model. We develop a variational EM framework for optimization where an explanation generation module and an explanation-augmented prediction module are alternatively optimized and mutually enhance each other. Moreover, we further propose an explanation-based self-training method under this framework for semi-supervised learning. It alternates between assigning pseudo-labels to unlabeled data and generating new explanations to iteratively improve each other. Experiments on two natural language understanding tasks demonstrate that our framework can not only make effective predictions in both supervised and semi-supervised settings, but also generate good natural language explanation.