어형적 연관성을 고려한 어휘 빈도 산출

Abstract

Jae-Woong Choe, Hyejoon Yoon, Jieun Jeon. 2011. 08. Extraction of Lexical Frequency based on Morphological Relatedness. Korean Linguistics 52, 243-272. This paper starts from the claim as in Han (2006) that morphological relatedness, or word family, should be taken into consideration when calculating word frequency in Korean. In this paper, a specific method, implemented in Perl, is proposed to gather Korean word family members and their frequency for a given root word. The target corpus used for extraction is the Sejong 21 sense tagged corpus. The resulting list shows some interesting points in terms of the effects of word family for frequency count, and also in terms of the comparison with the simple frequency count of the head words. Spearman’s Correlation Coefficient is applied to quantify the relationship between the head word frequency list and the word family frequency list, and it is observed that there is an extremely strong relationship (0.906) between the two.