Graph sentence neural networks can be employed to model cross-sentence dependencies for relation extraction, which bring in memory and reasoning abilities


Columbia University

- Syntactic methods
- Multi-hop paths between entities
- Graph-structured neural networks

Methods for handling more complicated context in RE

Syntactic methods rely on textual features extracted from varied syntactic structures like coreference annotations, dependency parsing trees, and discourse relations, to connect sentences in documents.


Syntactic methods for RE from complicated context 

Inter-sentence entity graphs can use multi-hop paths between entities for inferring the correct relations. 


Learn Before

Related