구글 트렌드를 활용한 일본어 유의어 분석 가능성 고찰 = A Study on the Analyzability of Japanese Synonyms Based on Google Trends

다국어 초록 (Multilingual Abstract)

In this study, I propose a new method of analysis based on Google Trends to distinguish synonyms in Japanese. In particular, I focus on the ‘related themes’ and ‘related search words’ of the Google Trends data and analyze three pairs of synonyms, ｢kanarazu｣ and ｢kitto｣ ‘for sure’, ｢kaizen｣ and ｢kairyou｣ ‘improvement’, and ｢daiji｣ and ｢taisetsu｣ ‘important’. Through the analysis, I reveal subtle meaning differences more clearly than the previous corpus analyses and the dictionary definitions. The results of this study are condensed into three points. First, kanarazu is related to themes such as effort, quotes, traffic signs, and prohibition, whereas kitto tends to co-occur with search words describing emotions such as hope and sorrow. Second, kaizen is strongly associated with the themes of health and beauty, such as hair, wrinkles, fat, liver, constipation, waist, and long-term care. By contrast, kairyou is related to the themes of agriculture and ground such as soil, soil improvement, breeding, and cultivated varieties. Finally, daiji is mainly searched for in connection with body-related themes, whereas taisetsu is searched for in connection with themes such as persons, personality, friendship, mind, and quotes.