RESEARCH ON PROOFREADING ALGORITHM OF MONGOLIAN HOMOGRAPH ERRORS BASED ON FINITE STATE AUTOMATA

摘要

For homograph errors in Mongolian texts, this paper integrates dictionary and wordformation rules into finite state automata model, and introduces a proofreading method of Mongolian homograph errors based on finite state automata. Experimental results indicate that this method can quickly proofread Mongolian texts and the average correction rate is 91.5%. This paper integrates dictionary and word-formation rules into finite state automata model, using the rule base of isomorphic character, performing heuristic search in the lexical analyzer to check error and correct error. This method is suitable for error check and error correction of non-word error, and is not suitable for error check and error correction of real-word, syntax and semantic error.