We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Large Scale Legal Text Classification Using Transformer Models

Abstract: Large multi-label text classification is a challenging Natural Language Processing (NLP) problem that is concerned with text classification for datasets with thousands of labels. We tackle this problem in the legal domain, where datasets, such as JRC-Acquis and EURLEX57K labeled with the EuroVoc vocabulary were created within the legal information systems of the European Union. The EuroVoc taxonomy includes around 7000 concepts. In this work, we study the performance of various recent transformer-based models in combination with strategies such as generative pretraining, gradual unfreezing and discriminative learning rates in order to reach competitive classification performance, and present new state-of-the-art results of 0.661 (F1) for JRC-Acquis and 0.754 for EURLEX57K. Furthermore, we quantify the impact of individual steps, such as language model fine-tuning or gradual unfreezing in an ablation study, and provide reference dataset splits created with an iterative stratification algorithm.
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as: arXiv:2010.12871 [cs.CL]
  (or arXiv:2010.12871v1 [cs.CL] for this version)

Submission history

From: Zein Shaheen [view email]
[v1] Sat, 24 Oct 2020 11:03:01 GMT (954kb,D)

Link back to: arXiv, form interface, contact.