本文旨在探討顧客在購買圖書時的特性及行為,以作為出版商在通路、擺放、綁售等行銷策略擬定時的參考,以提高圖書銷售營業額。本研究收集某雜誌出版業者自2010年下半年的顧客交易資料,透過資料挖掘中的關聯規則技術來探索圖書消費市場潛在的特性,透過購物籃(market basket)分析概念,用以瞭解顧客在購買圖書時,不同屬性出版品間可能的潛在關聯性,作為未來交叉銷售實務上有效建議。研究結果顯示,本次研究透過購物籃分析篩選出最佳的十組組合,包括3C數位商品與財經雜誌的組合、生活家電用品與養生雜誌組合等,出版商可以最佳組合搭配特殊節日或節氣、新聞議題、熱門與冷門產品、每個區域的消費習性等方式交叉銷售或擺放,以達到提昇銷售目的。
Purpose of this study was to explore customers purchase publications behavioral and relations for selling strategy. The item association in market basket by using association information clustering method, and understand the customer attribute when purchase publications, thus complete the classification of data mining, as practical recommendations for cross-selling. In this study, customer transaction data was collected from a magazine publishing company during the second half of 2010. The results of the research show that 10 optimal combinations was identified by market basket analysis, publishers can use these best combinations with special holidays, news topics, hot and cold products, the consumption habits of each region to make cross-sale or display, etc., in order to improve sales.