In-depth analysis of the impact of OCR errors on named entity recognition and linking.
Ahmed HamdiElvys Linhares PontesNicolas SidereMickaël CoustatyAntoine DoucetPublished in: Nat. Lang. Eng. (2023)
Keyphrases
- named entity recognition
- information extraction
- named entities
- recognition errors
- natural language processing
- maximum entropy
- text summarization
- semi supervised
- conditional random fields
- sequence labeling
- document images
- optical character recognition
- hand coded
- error correction
- relation extraction
- maximum entropy classifier
- annotated corpus
- information retrieval
- question answering
- classifier ensemble
- object recognition
- proper names
- data sets