Legal entity recognition in indonesian court decision documents using Bi-LSTM and CRF approaches

Eka Qadri Nuranti, Evi Yulianti

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Citations (Scopus)

Abstract

The increasing number of Indonesian court decision documents requires special attention to be managed and used efficiently. One thing is to recognize legal entities to identify relevant legal information in the documents. This research examines the effectiveness of several deep learning methods, to recognize ten legal entities in Indonesian court decision documents. In this task, we found that the combination of Bi-LSTM and CRF methods achieve the highest F-1 score of 0.83. It outperforms other deep learning methods (CNN, LSTM, LSTM+CRF and Bi-LSTM) by 2 - 12 % and machine learning methods (SVM and CRF) by 41 - 76 %. The results show that this model can be used to identify relevant information about the legal domain in Indonesia.

Original languageEnglish
Title of host publication2020 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages429-434
Number of pages6
ISBN (Electronic)9781728192796
DOIs
Publication statusPublished - 17 Oct 2020
Event12th International Conference on Advanced Computer Science and Information Systems, ICACSIS 2020 - Virtual, Depok, Indonesia
Duration: 17 Oct 202018 Oct 2020

Publication series

Name2020 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2020

Conference

Conference12th International Conference on Advanced Computer Science and Information Systems, ICACSIS 2020
Country/TerritoryIndonesia
CityVirtual, Depok
Period17/10/2018/10/20

Keywords

  • Bi-lstm
  • CRF
  • Legal document
  • Legal entity recognition
  • Legal processing
  • Lstm
  • Name entity recognition
  • ner

Fingerprint

Dive into the research topics of 'Legal entity recognition in indonesian court decision documents using Bi-LSTM and CRF approaches'. Together they form a unique fingerprint.

Cite this