# Interpretable Structure-aware Document Encoders with Hierarchical Attention

## TL;DR

This paper introduces a hierarchical, structure-aware document encoder called Structure Tree-LSTM that improves document classification by capturing internal structure and providing interpretability through attention visualization.

## Contribution

It extends Tree-LSTMs with hierarchical attention to better model larger texts and demonstrates improved performance and interpretability over baselines.

## Key findings

- Outperforms baselines on two datasets using structural clues.
- Maintains performance with both sentence and word embeddings.
- Achieves competitive results on a medical report dataset.

## Abstract

We propose a method to create document representations that reflect their internal structure. We modify Tree-LSTMs to hierarchically merge basic elements such as words and sentences into blocks of increasing complexity. Our Structure Tree-LSTM implements a hierarchical attention mechanism over individual components and combinations thereof. We thus emphasize the usefulness of Tree-LSTMs for texts larger than a sentence. We show that structure-aware encoders can be used to improve the performance of document classification. We demonstrate that our method is resilient to changes to the basic building blocks, as it performs well with both sentence and word embeddings. The Structure Tree-LSTM outperforms all the baselines on two datasets by leveraging structural clues. We show our model's interpretability by visualizing how our model distributes attention inside a document. On a third dataset from the medical domain, our model achieves competitive performance with the state of the art. This result shows the Structure Tree-LSTM can leverage dependency relations other than text structure, such as a set of reports on the same patient.