Research on Mongolian lexical analyzer based on NFA

Mongolian is an adhesive language. Its word-formation and configuration is built through the stem is connected to different suffixes. In theory, Mongolian vocabulary is unlimited, so the dictionary can not encompass all of the words and their numerous morphological changes. Development of independent, efficient lexical analyzing software to identify and generate the words and their morphological changes is needed. In this paper, we have introduced a Mongolian lexical analyzer, which has used dictionaries and NFA-based methods to greatly improve the speed of analyzing. After used in the modern Mongolian parsing software, we found that compare with the simple dictionary or rules-based algorithm it improves the speed by nearly two orders of magnitudes.