長期以來，日籍旅客為主要來台外籍旅客，未來如何維持，甚至提升日籍旅客數，進而拉近觀光貿差，為觀光單位需關注之問題。本研究針對觀光局2004年及2014年「來台旅客消費及動向調查」之日籍旅客資料，以大數據觀點分析來台日籍旅客特性，並應用停留天數、購買金額、購買金額為旅客價值區隔變數，採用C5.0決策樹做為資料探勘技術，比較分析兩年度高價值日籍顧客屬性特徵與消費行為，提供觀光策略制定者與業者於行銷之參考。

For a long time, Japanese tourists have constituted the majority of international tourism in Taiwan. To maintain their level of interest and even increase the scale of inbound tourism from Japan, thus reducing tourism deficit, is an issue that concerns the Tourism Bureau. This study conducted big data analysis based on the ＂Annual Survey of Visitors Expenditure and Trends in Taiwan＂ by the Tourism Bureau of the Republic of China in 2004 and 2014 to investigate the attributes of inbound visitors from Japan. Length of stay and expenditure were used as segmentation variables for visitor value with the C5.0 decision tree data mining method. The different attributes and consumer behavior of Japanese visitors from the two years were compared for reference in policymaking in the public sector and marketing in the private sector.