DocRED: A Large-Scale Document-Level Relation Extraction Dataset.
Yuan YaoDeming YePeng LiXu HanYankai LinZhenghao LiuZhiyuan LiuLixin HuangJie ZhouMaosong SunPublished in: CoRR (2019)
Keyphrases
- relation extraction
- document level
- coreference resolution
- automatic extraction
- information extraction
- domain specific
- semantic relations
- named entities
- query expansion
- semantic features
- sentence level
- question answering
- language model
- sentiment classification
- named entity recognition
- natural language processing
- pseudo relevance feedback
- conditional random fields
- web pages
- hidden variables
- relevance feedback
- semi supervised
- knowledge discovery
- similarity measure