Unsupervised Multi-View Post-OCR Error Correction With Language Models.
Harsh GuptaLuciano Del CorroSamuel BroscheitJohannes HoffartEliot BrennerPublished in: EMNLP (1) (2021)
Keyphrases
- error correction
- multi view
- language model
- semi supervised
- language modeling
- single view
- multiple views
- n gram
- probabilistic model
- d objects
- depth map
- information retrieval
- three dimensional
- unsupervised learning
- mixture model
- supervised learning
- error detection
- co training
- error control
- unlabeled data
- watermarking scheme
- semi supervised learning
- labeled data
- smoothing methods
- training set