국경 간 전자상거래 소셜 데이터 마이닝을 활용한 마케팅 컨텐츠 연구: 중국 온라인 플랫폼의 한국화장품 구매후기를 예로

Abstract

With the development of the online industry, reviews of platform use play an important role in obtaining product information from cross-border e-commerce users and affect purchase intentions. In this context, text analysis of purchase review data was performed to derive consumer emotional factors. The data for analysis were targeted at purchase reviews for Korean cosmetics sold on Chinese platforms. Word cloud and emotional analysis were performed on the collected text data by deriving words with high frequency through a preprocessing process. For the analysis, response variables were applied to each label in the late purchase, and the TF-IDF algorithm value was used as an input variable. In addition, through vector streaming based on the evaluation SD-SVM trust section, inducement factors that express emotions such as positive, neutral, and negative after purchasing and using the target product were derived. As a result of the analysis, positive factors such as freshness, purity, functionality, and neutrality of consumers' emotions were highlighted, and negative factors such as service attitude, price, quality, and fake goods were also positiveIt appeared at a similar frequency to that of cattle. Therefore, the results of the study show that the positive factors are further strengthened to increase the competitiveness of the product, Neutral and negative factors which can be seen as a phenomenon in which consumer interest decreases, can be provided as evidence that requires efforts to remove and improve. It is also meaningful in that it can be used as marketing content for the local online market by visualizing the emotional expression of cross-border platform users.