A Novel Method for Open Relation Extraction from Public Announcements of Chinese Listed Companies
Tianyuan Zhuang, Pengwei Wang, Yaying Zhang, Zhaohui Zhang · 2017
There exists massive information on the Internet, consequently information extraction has attracted more and more attention. Open relation extraction aims to get relational tuples from text without defining relation types in advance. This paper focuses on identifying relation phrases and associated arguments in Chinese arbitrary sentences, and proposes a domain-adaptive method to get entity relations from large amount of related Chinese texts in a supervised way. The experiments show that the proposed method can get both good accuracy and recall rate in the domain of Chinese listed companies.