This work is part of a more general project aiming to design a tool that can help lawyers to find the information they need for litigation in a fast and efficient way. The resource is being designed for Spanish, a language that has a scarceness of Natural Language applications for legal coding, and is tested in 300 documents, mainly writs of ‘amparo’, a legal procedure to protect human rights, by means a judicial review of governmental action. These documents have been freely downloaded from the Mexican Instituto Federal de Telecomunicaciones. The system, implemented in Python, will include modules to perform several tasks, like automatic classification, Named Entities identification, law detection, structure summarization, and event extraction. This article is focused in one of the most complex parts of the development, event extraction. The algorithm works linking dates with events in the texts. These events are reduced to a list of verbs that have been reported as the most meaningful in this type of texts. For every verb-event, a list of pieces of information will be retrieved: ‘who’, ‘what’, ‘to whom’ and ‘where’.
@InProceedings{SIERRA18.8, author = {Gerardo Sierra ,Gemma Bel-Enguix ,Guillermo López-Velarde ,Ricardo Saucedo and Lucía Rivera}, title = {Event Extraction from Legal Documents in Spanish }, booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)}, year = {2018}, month = {may}, date = {7-12}, location = {Miyazaki, Japan}, editor = {Georg Rehm and Víctor Rodríguez-Doncel and Julián Moreno-Schneider}, publisher = {European Language Resources Association (ELRA)}, address = {Paris, France}, isbn = {979-10-95546-18-4}, language = {english} }
Summary of the paper