Date: Oct.10^th, 2019

Time: 18:30-20:00

Speaker: LIU Xiaozhong

Venue: Room 250, Mong Man Wai Building, Zijingang Campus

【Speaker Intro】

Xiaozhong Liu is Associate Professor from School of Informatics, Computing and Engineering at Indiana University Bloomington. His research is focusing on Information Retrieval, NLP, Digital Library, Information Seeking and Heterogeneous Graph Mining. While active in academia, his research also extends to industry. Currently, his algorithms are used more than 50 million times per day for a number of world-leading search engines and recommendation systems.

【Abstract】

The success of machine (deep) learning amazes the audience from various domains, e.g., recent AI efforts transcend humanity with cognitive awareness for a number of predefined tasks. Legal domain AI, however, can be more challenging and bewildering than other text mining/NLP disciplines. In this talk, I will investigate a number of novel legal tasks, i.e., Cybercrime Detection, eCommerce Dispute Legal Prediction, and Legal Summarization for Court Debate Dialog, by leveraging innovative natural language processing and deep learning methodologies. Unlike prior efforts in machine learning, legal knowledge base and law-reasoning can play an essential role for legal AI tasks. The complex relations among the concepts in a particular law domain, e.g. intellectual property or personal injury liability, can be an important prerequisite for case representation learning. Meanwhile, a legal case should be a heterogeneous and multi-view instance for mining, and different components, e.g., plaintiff, defendant, facts, undisputed legal concepts, precedents, and controversy focuses, should play different roles for a comprehensive case legal characterization.