Categorical Ambiguity and Information Content: A Corpus-based Study of Chinese

The degree of ambiguity in Chinese is investigated in this paper based on the tagged Sinica Corpus. We propose to use measurement of information content instead of frequency to model generalizations about categorical ambiguity Two important generalizations were found: First, the degree of ambiguity indeed correlates with the number of possible categories that a word has. Second,. Even though frequently used words are more likely to be categorical ambiguous (Huang et al. 2002), the degree of ambiguity of a word does not depend on its frequency.