Extreme Multi-Label Legal Text Classification: A case study in EU Legislation
Ilias Chalkidis* Manos Fergadiotis* Prodromos Malakasiotis* Nikolaos Aletras** Ion Androutsopoulos* * Department of Informatics, Athens University of Economics and Business, Greece ** Computer Science Department, University of Sheffield, UK
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 eur-lex, the European Union’s public document database, annotated with concepts from eurovoc, a multidisciplinary thesaurus. The dataset is substantially larger than previous eur-lex 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.
中文速览
给欧盟立法文件自动打标签是个棘手的任务，因为标签体系（EuroVoc叙词表）有逾7000个概念，且大量标签在实际文件中极少出现。研究者为此发布了一个全新大规模数据集EurLex57K，收录5.7万份欧盟立法英文文件，比此前同类数据集大出一倍以上，同时天然支持小样本和零样本学习场景。他们在这个数据集上系统比较了多种神经网络分类器，发现使用自注意力机制的双向门控循环单元（BiGRU with self-attention）显著优于当前主流的标签级注意力网络（Label-Wise Attention Network，LWAN），而把LWAN中的CNN编码器换成BiGRU后性能还能进一步提升，在零样本场景下同样如此。这项工作不仅填补了法律NLP领域高质量公开基准的空白，也纠正了此前"标签级注意力一定优于通用序列模型"的认知偏差，为后续研究提供了更扎实的数据和方法基准。
原文 arXiv:1905.10892;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1905.10892v1