隨著網路購物的普及，對於電子商務平台業者來說，如何精準確認消費者的瀏覽意圖是很重要的。若能將消費者每次的瀏覽行為視為一單位，而非將同一位消費者所有的瀏覽行為囊括為一單位來進行分析，就能避免同一位消費者的瀏覽意圖在不同階段或時期可能會不同，進而造成結果不精確的狀況發生。然而，過去的研究鮮少將此因素納入考量。本篇論文以3C購物網站的用戶瀏覽行為進行分析，以每次的瀏覽行為為單位來進行消費者瀏覽意圖研究，並著重在頁面停留時間百分比變數，透過分群將消費者的瀏覽行為模式分類，接著用決策樹分析了解每個群集的特徵，並依特徵將各群集配對顧客旅程的各階段，最後給予處於各階段的群集相關的行銷建議。;With the popularization of online shopping, e-commerce platform companies need to define customers’ browsing intentions precisely. If we can make every customer’s each browsing session but every customer’s all browsing history as a unit, we can avoid the situation in which one customer has different browsing intentions in different periods and make the research results not be influenced by this deviation. Nevertheless, previous researches rarely consider this component. This study aims to take every customer’s each browsing session as a unit to analyze customers’ browsing intentions and focus on page browsing time proportion. First, it classifies customers’ browsing behaviors into four clusters by K-means Clustering. Second, it uses Decision trees to analyze each cluster’s attributes and map these clusters to different Customer journey stages. Finally, it gives marketing suggestions for each cluster based on its attributes and the customer journey stage’s meaning.
