作者: Sargur N. Srihari , Jonathan J. Hull , Ramesh Choudhari
关键词: Computational complexity theory 、 Artificial intelligence 、 Representation (mathematics) 、 Speech recognition 、 Computer science 、 Trie 、 Error detection and correction 、 Soft output Viterbi algorithm 、 Natural language processing 、 Viterbi algorithm 、 Character (mathematics) 、 String (computer science)
摘要: This paper presents an efficient method for the integration of two forms contextual knowledge into correction character substitution errors in words text: bottom-up form transitional probabilities and top-down a dictionary. The is modification Viterbi algorithm---which maximizes string posteriori probability by using confusion probabilities---so that only legal strings are output. algorithm achieves its efficiency trie structure representation dictionary search process. An analysis computational complexity results experimentation with approach presented.