A Data-driven Approach for Noise Reduction in Distantly Supervised Biomedical Relation Extraction.
Saadullah AminKatherine Ann DunfieldAnna VechkaevaGuenter NeumannPublished in: BioNLP (2020)
Keyphrases
- noise reduction
- relation extraction
- information extraction
- entity extraction
- weakly supervised
- manually annotated
- automatic extraction
- named entities
- signal to noise ratio
- semantic relations
- domain specific
- named entity recognition
- semantic features
- noisy environments
- text mining
- semi supervised
- semantic role labeling
- edge detection
- question answering
- supervised learning
- broad coverage
- learning algorithm
- unsupervised learning
- natural language processing
- structured data
- co occurrence
- tree kernels
- machine learning
- dependency trees
- low level
- annotated corpus
- image classification
- feature selection
- computer vision