@inproceedings{b926ef13fec8430a93fd87bb498a40f9,
title = "Event extraction for criminal legal text",
abstract = "This paper concerns with the actual problems in the legal work. We apply event extraction technology to the case description part in the Chinese legal text. We define the event type, event argument and event argument role of the larceny case, and construct a larceny case event extraction dataset through data annotation. We divide event extraction into two steps: event trigger word and argument joint extraction and event argument role assignment. We use BERT to obtain Chinese character vectors, use the BiLSTM-CRF model for extraction at the first step, and combine additional features with the extraction results of the first step, then input them to the CRF model of the second step to obtain an improvement in extraction result. We display the extracted event information in time series to realize the litigation visualization. We format Chinese time expressions, sorts the event information in tine series, and develops a Web application to display the timeline of event information.",
keywords = "Chinese legal text, Event dataset construction, Event extraction, Litigation visualization",
author = "Qingquan Li and Qifan Zhang and Junjie Yao and Yingjie Zhang",
note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 11th IEEE International Conference on Knowledge Graph, ICKG 2020 ; Conference date: 09-08-2020 Through 11-08-2020",
year = "2020",
month = aug,
doi = "10.1109/ICBK50248.2020.00086",
language = "英语",
series = "Proceedings - 11th IEEE International Conference on Knowledge Graph, ICKG 2020",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "573--580",
editor = "Enhong Chen and Grigoris Antoniou and Xindong Wu and Vipin Kumar",
booktitle = "Proceedings - 11th IEEE International Conference on Knowledge Graph, ICKG 2020",
address = "美国",
}
