Character-based Bidirectional LSTM-CRF with words and characters for Japanese Named Entity Recognition.
Shotaro MisawaMotoki TaniguchiYasuhide MiuraTomoko OhkumaPublished in: SWCN@EMNLP (2017)
Keyphrases
- named entity recognition
- chinese characters
- conditional random fields
- information extraction
- character recognition
- named entities
- natural language processing
- maximum entropy
- sequence labeling
- handwritten words
- proper names
- semi supervised
- graphical models
- handwritten characters
- writing styles
- text summarization
- printed characters
- relation extraction
- text lines
- crf model
- printed documents
- probabilistic model
- hidden markov models
- higher order
- annotated corpus
- optical character recognition
- text documents
- text mining
- n gram
- markov random field
- noun phrases
- classifier ensemble
- word recognition
- document images
- generative model
- pos tagging
- co occurrence
- pairwise
- maximum entropy classifier