Extreme Multi-Label Legal Text Classification: A case study in EU Legislation

TL;DR

This paper introduces a large-scale legal text dataset for extreme multi-label classification and demonstrates that BIGRU-based models with self-attention outperform existing methods in this domain.

Contribution

It provides a new, extensive EURLEX dataset suitable for XMTC, few-shot, and zero-shot learning, and shows that BIGRU models with self-attention outperform prior state-of-the-art methods.

Findings

BIGRU with self-attention outperforms label-wise attention models

Replacing CNNs with BIGRUs improves performance

The dataset enables better evaluation of XMTC in legal texts

Abstract

We consider the task of Extreme Multi-Label Text Classification (XMTC) in the legal domain. We release a new dataset of 57k legislative documents from EURLEX, the European Union's public document database, annotated with concepts from EUROVOC, a multidisciplinary thesaurus. The dataset is substantially larger than previous EURLEX datasets and suitable for XMTC, few-shot and zero-shot learning. Experimenting with several neural classifiers, we show that BIGRUs with self-attention outperform the current multi-label state-of-the-art methods, which employ label-wise attention. Replacing CNNs with BIGRUs in label-wise attention networks leads to the best overall performance.