ACL RD-TEC 1.0 Summarization of P00-1037
Paper Title:
AN IMPROVED ERROR MODEL FOR NOISY CHANNEL SPELLING CORRECTION
AN IMPROVED ERROR MODEL FOR NOISY CHANNEL SPELLING CORRECTION
Authors: Eric Brill and Robert C. Moore
Primarily assigned technology terms:
- algorithm
- character insertion
- computing
- damerau-levenshtein distance
- dynamic programming
- dynamic programming algorithm
- error reduction
- expectation-maximization
- expectation-maximization algorithm
- learning
- levenshtein
- modeling
- noisy channel model
- noisy channel spelling
- parameter reestimation
- partitioning
- programming algorithm
- reestimation
- rescoring
- search
- search technique
- searching
- spell checker
- spell checking
- spelling
- spelling correction
Other assigned terms:
- approach
- brown corpus
- case
- characters
- checker
- cognitive
- context window
- contextual information
- dictionary
- distribution
- edit distance
- error rate
- feature
- input string
- language model
- language models
- meaning
- measure
- measures
- morpheme
- morpheme boundary
- n-best list
- noise
- noisy channel
- probabilities
- probability
- process
- relative error reduction
- segments
- technique
- term
- test set
- text
- text corpus
- training
- training data
- training set
- transposition
- trigram
- trigram language model
- trigram model
- uniform probability
- weight vector
- window size
- word
- word corpus
- words
