近代電子商務的快速成長,尤其宅世代的來臨,使大眾由一般購物轉向網路購物,在網路的大量的搜尋和比價的風氣盛行下,企業必須使用大量的促銷以吸引顧客,而企業藉由了解每位顧客特色與不同需求,分析顧客的購買行為與預測其喜好,能減少不當的成本,提供適當的產品、促銷與服務給顧客。 本研究考量「顧客基本資料」、「顧客搜尋行為」及「購買商品價格」,使用粒子群最佳化演算法(Particle Swarm Optimization, PSO)將顧客區分為不同特性的群集,分析各群集對於折價、贈品及折價券的喜好,找出顧客適合的促銷策略,並藉由顧客對促銷滿意度來判定顧客喜愛的促銷方法。 本研究發現不同顧客搜尋行為會影響購買行為,藉由資料探勘找出不同的顧客行為會選擇不同的促銷策略,而經由促銷滿意度找出顧客喜愛促銷策略,使得在企業更能提供更適切的促銷策略給予消費者,減少資源上的浪費且能降低企業成本。
The electronic commerce grows fast in modern times. Especially the coming of new generations, the people change from physical store shopping to online shopping. A large number of search from Internet and the price relation is frequent. Companies must use a large number of promotions to attract customers. The Companies try to understand the characteristics of each customer with different needs, and analyze the buying behavior of customers and predict their preferences, it can reduce the inappropriate cost, and the provision of appropriate products, promotions and services to customers. This research find customer profile form "customer information", "customer search behavior" and "purchase price". In this study of particle swarm optimization algorithm (PSO) is developed to differentiate the customers for different characteristic clusters. The clustering Technique in applied to analyze the preferences from discount, gift and coupons. Based the lustered result, the appropriate promotion strategy is determined to satisfy customer demand. This research found that search behavior of different customers will affect their buying behavior. Then data mining is used to identify the different behaviors of customers and choose the different promotion strategy. Through the process, it can find favorite promotion strategy of customer. Companies can provide a more appropriate promotion strategy to consumers, and reduce the waste of resources and cost.