本論文在探討分析序列型樣其間隔天數的銷售量之趨勢以設計一套推薦系統。經由統計產品銷售紀錄,我們以間隔天數與銷售量可以得到一個波形圖,此波形圖橫軸為間隔天數縱軸為銷售量,經由趨勢偵測可得知波形的起伏並得以瞭解消費者的購買習性,了解什麼時候可以推薦什麼產品給消費者以刺激她們的購買慾望,並期望能減少消費者購買產品之間的間隔天數,相對的亦能夠促使銷售量上升,且針對重點的消費族群和產品做推銷亦能降低成本花費,如此可以提供決策者一個不錯的建議。 This study confers on the trend of each day-interval’s sales volume of sequential pattern to design a recommend system. By accounting the selling records of products, we can get an oscillograph by day-interval and sales volume, this oscillograph sets the interval days as the horizontal axle and sales volume as the vertical axle. Depending on trend detection, we can know that the trend of the wave goes up or down, and we will understand the shopping habits of consumers. This can help us to know when to recommend products to customers, and hope that they will shorten the interval of days between buying products, relatively raising the sales volume. Furthermore, sales promotion on the important product toward the important consumers can reduce the production cost, so we can provide a good suggest for policymaker.