Class-Adaptive Self-Training for Relation Extraction with Incompletely Annotated Training Data.
Qingyu TanLu XuLidong BingHwee Tou NgPublished in: ACL (Findings) (2023)
Keyphrases
- relation extraction
- manually annotated
- training data
- annotated corpus
- information extraction
- automatic extraction
- training set
- named entities
- domain specific
- question answering
- semantic role labeling
- semantic relations
- data sets
- semantic features
- dependency trees
- classification accuracy
- learning algorithm
- named entity recognition
- text classification
- parse tree
- broad coverage
- tree kernels
- supervised learning
- support vector machine
- data points
- domain knowledge
- decision trees
- knowledge base
- data mining