Unsupervised Learning of Explainable Parse Trees for Improved Generalisation

TL;DR

This paper introduces an attention-based Tree-LSTM model that learns more interpretable and meaningful parse trees, leading to improved performance across various NLP tasks and better linguistic structure discovery.

Contribution

It proposes a novel attention mechanism over Tree-LSTMs for learning explainable parse trees, enhancing interpretability and semantic correctness in NLP representations.

Findings

Improved performance on natural language inference, semantic relatedness, and sentiment analysis.

Learned parse trees are more explainable and linguistically meaningful.

Model outperforms recent RvNN-based methods.

Abstract

Recursive neural networks (RvNN) have been shown useful for learning sentence representations and helped achieve competitive performance on several natural language inference tasks. However, recent RvNN-based models fail to learn simple grammar and meaningful semantics in their intermediate tree representation. In this work, we propose an attention mechanism over Tree-LSTMs to learn more meaningful and explainable parse tree structures. We also demonstrate the superior performance of our proposed model on natural language inference, semantic relatedness, and sentiment analysis tasks and compare them with other state-of-the-art RvNN based methods. Further, we present a detailed qualitative and quantitative analysis of the learned parse trees and show that the discovered linguistic structures are more explainable, semantically meaningful, and grammatically correct than recent approaches. The source code of the paper is available at https://github.com/atul04/Explainable-Latent-Structures-Using-Attention.