Mining Entity-Relation-Attribute Structures from Massive Text Data
Nov 21, 2017 1161 views 
Entity-Relation-Attribute (ERA) structures, forming structured networks between entities and attributes, have demonstrated the flexibility of storing rich information and the effectiveness of gaining insights and knowledge. However, the majority of massive amount of data in the real world are unstructured text, ranging from news articles, social media post, advertisements, to a wide range of textual information from various domains (medical records, corporate reports). Without heavy human annotations and curations, most of existing approaches have difficulties in extracting named entities and their relations as well as typing and organizing knowledge as networks. Link to tutorial: https://shangjingbo1226.github.io/2017-08-11-kdd-tutorial/
