????? ¡®ÉËÐÄ¡¯? ??????? ?????? ?? ?? -??? ??? BP??? (A Study on the Semantic Network and Differences of Synonymous Emotional Adjectives Meaning ¡®SAD¡¯ in Modern Chinese: Corpus Based BP Approach)

? ??? ??? ¡®ÉËÐÄ¡¯, ¡®ÄÑÊÜ¡¯, ¡®Äѹý¡¯, ¡®Í´¿à¡¯ ??? ?? ??? ??? ?? BP???? ??? ??? ???? ¡®?? ???¡¯? ???? ?? ??? ??? ??? ??? ??? ??? ?? ??? ???? ?????. ??? ??? ???? ?? CA??? HAC??? ?? ?? ??? ??? ???? ???? ??? ?????.
??? BP???? ?? ???? ?? ¡®Äѹý¡¯? ¡®ÄÑÊÜ¡¯? ?? ???? ?? ?? ?????? ??? ?? ¡®ÉËÐÄ¡¯? ????, ¡®Í´¿à¡¯? ??? ?? ??? ???. ??? ??? ????? ¡®Í´¿à¡¯? ??? ??(??? ??, ???, ???? ???) ??? ??? ??? ???, ¡®Äѹý¡¯? ¡®ÄÑÊÜ¡¯? ??? ????? ????? ?? ??, ??? ¡®ÉËÐÄ¡¯? ¡®Í´¿à¡¯? ???? ????? ?? ??? ??? ??? ?????. ?? ??? ??? ??? ?? ??? ??? ??? ?? ??? ???? ¡®ÄÑÊÜ¡¯? ¡®Äѹý¡¯? ????? ?? ??? ??? ????. ??? ??? ?? ???? ?? ¡®ÄÑÊÜ¡¯? ¡®Äѹý¡¯? ??? ???? ?????. ????? ??? ???? ?? ??? ??? ¡®ÄÑÊÜ¡¯? ¡®Äѹý¡¯? ?? ??, ? ??? ¡®ÉËÐÄ¡¯? ??, ??? ¡®ÉËÐÄ¡¯? ¡®Í´¿à¡¯? ?? ??? ??? ??? ?????.

This study investigated the ¡®behavioral profile' in corpus use cases by applying the BP approach to reveal the difference in meaning between the synonyms ¡®shangxin', ¡®nanshou', ¡®nanguo', and ¡®tongku' groups. Through this method, integrated data was derived on the characteristics and differences of syntactic functions and various semantic dimensions. And on these data, CA analysis and HAC analysis were performed to explore the structure within the category, and the structure of the network was proposed.
As a result of cluster analysis on the syntactic BP data, ¡®nanguo' and ¡®nanshou' were the most similar and clustered first, and this cluster was combined with ¡®shangxin' again, and ¡®tongku' was the most distant from them. These differences are specifically related to the preference for the noun use of ¡®tongku' (noun subject, object, and the object of the preposition) and the non-preference for the descriptive use, the preference for the use of the complement and predicate of ¡®nanguo' and ¡®nanshou' as predicates in the predicate structure, and It was analyzed that the preference for using the adverbs and adjectives of ¡®shangxin' and ¡®tongku' was caused. Clustering analysis of all semantic dimension features suggested that ¡®nanshou' and ¡®nanguo' are internally closest similar to syntactic distribution characteristics. For the integrated profile of syntactic and semantics, both distance measures ¡®nanshou' and ¡®nanguo' were proposed as one cluster. And ¡®shangxin' was similar to these and ¡®tongku' was found to be the most different from the other three words. Finally, based on the proposal for this network, the difference in meaning between ¡®nanshou¡¯ and ¡®nanguo¡¯, the difference between this cluster and ¡®shangxin¡¯, and the main difference and the typical meaning between ¡®shangxin¡¯ and ¡®tongku¡¯ were explained.