Analysis of fashion market trend using advertising data of shopping information site

Abstract

In recent years, consumer needs for products have been diversified, and companies need proper site management considering trends. As the needs of consumers diversified, it became difficult to grasp the needs. In this study, we analyzed diversifying customer needs by focusing on internet advertising. We classify it into a company’s products by Doc2Vec using each shopping site documents to segment the market using word similarity in the fashion market and verify that a company can accurately identify the needs of consumers in fashion sites. Afterwards, we consider the impression tendency of sites with similar ones by using time-series clustering.