On the automatic analysis of rules governing online communities

Adan Beltran*, Nardine Osman, Lourdes Aguilar, Carles Sierra

*Corresponding author for this work

Research output: Chapter in BookChapterResearchpeer-review

Abstract

The automatic translation of rules or legal text from natural language into formal language has gained interest in the natural language processing domain, especially in the field of law and AI. Our research goal is to be able to automatically extract, from rules in natural language, the necessary elements that define these rules, such as the action in question, its modality (duty, right, privilege,..), the first person the rule addresses, the second person affected by the rule, and the condition (if the rule was a conditional rule). As a first step toward identifying these elements, we start by identifying the semantic subjects, verbs, and objects in sentences of online normative texts. This paper presents the SVO+ model that achieves this, and our evaluation illustrates the model’s high precision when tested with the terms of use from websites like Facebook and Twitter.

Original languageAmerican English
Title of host publicationAdvances in Artificial Intelligence – IBERAMIA 2018 - 16th Ibero-American Conference on AI, Proceedings
EditorsEduardo Fermé, Guillermo R. Simari, Flabio Gutiérrez Segura, José Antonio Rodríguez Melquiades
Pages354-366
Number of pages13
DOIs
Publication statusPublished - 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11238 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Keywords

  • Normative texts
  • Objects
  • Semantic subjects
  • Text analysis
  • Verbs

Fingerprint Dive into the research topics of 'On the automatic analysis of rules governing online communities'. Together they form a unique fingerprint.

Cite this