Improving LSTM CRFs using character-based compositions for Korean named entity recognition.
Seung-Hoon NaHyun KimJinwoo MinKangil KimPublished in: Comput. Speech Lang. (2019)
Keyphrases
- named entity recognition
- information extraction
- conditional random fields
- named entities
- natural language processing
- maximum entropy
- sequence labeling
- semi supervised
- relation extraction
- recurrent neural networks
- text summarization
- classifier ensemble
- annotated corpus
- markov random field
- probabilistic model
- proper names
- named entity recognizer
- image processing
- question answering
- higher order
- co occurrence
- natural language
- training data