ACL RD-TEC 1.0 Summarization of P98-2138
Paper Title:
COMBINING TRIGRAM AND WINNOW IN THAI OCR ERROR CORRECTION
COMBINING TRIGRAM AND WINNOW IN THAI OCR ERROR CORRECTION
Authors: Surapant Meknavin and Boonserm Kijsirikul and Ananlada Chotimongkol and Cholwich Nuttee
Primarily assigned technology terms:
- algorithm
- approximation
- candidate generation
- character recognition
- contextsensitive spelling correction
- disambiguation
- dynamic programming
- dynamic programming technique
- error correction
- finite-state recognition
- hybrid method
- hypothesizing
- incremental algorithm
- information retrieval
- information retrieval system
- learning
- ocr error correction
- office automation
- optical character recognition
- programming technique
- real-word error correction
- recognition
- recognition algorithm
- retrieval system
- search
- segmentation
- segmentation algorithm
- speech recognition
- spelling
- spelling correction
- text recognition
- thai ocr
- thai ocr error correction
- voting
- weight updating
- word segmentation
Other assigned terms:
- ambiguity
- approach
- boundary ambiguity
- case
- characters
- collocation
- concept
- context words
- dictionary
- edit distance
- explicit word boundary
- fact
- feature
- feature-based approach
- generation
- hypotheses
- language model
- large corpus
- lattice
- method
- n-gram
- n-gram table
- opinions
- part of speech
- part-of-speech
- part-of-speech tags
- part-of-speech trigram
- part-of-speech trigram model
- phrase
- probabilities
- probability
- process
- search space
- sentence
- sentences
- sources of information
- spelling error
- statistical information
- substring
- tags
- target word
- technique
- terms
- test set
- text
- thai ocr error
- time complexity
- training
- training set
- trigram
- trigram model
- unknown word model
- word
- word boundary
- word boundary ambiguity
- word collocation
- word lattice
- word model
- word sequence
- word sequences
- words
- writing system
