Probability Distribution of Dependency Distance of Mongolian Nouns

Abstract

Based on the written and spoken language treebanks, this paper applies quantitative linguistics methods to statistically analyze Mongolian nouns and their dependency distance distributions. The results showed that the proportion of nouns in the Mongolian written language treebank was 35.58%, and the proportion of nouns and pronouns was 39.61%. In the Mongolian spoken language treebank, the proportion of nouns was 22.11%, and the proportion of nouns and pronouns was 33.12%. These findings closely align with results from prior research, further confirming the regularity of the proportion of nouns in human natural language. In terms of dependency distance distributions, the distributions of noun dependency distances in both the written and spoken treebanks fit the right-truncated Waring distribution and right-truncated modified Zipf-Alekseev distribution. This result is consistent with previous research. Moreover, the “long tail effect” can be seen from the fitting results, that is, as the dependency distance increases, its frequency decreases. This implies that there is also a tendency to minimize dependency distances in nouns.