| Summary: | 碩士 === 朝陽科技大學 === 資訊管理系碩士班 === 96 === In the competing environment of e-commerce, how merchant maintains customer’s loyalty and understand customer’s purchasing behavior has become the key factors for competition. In order to provide appropriate service to increase profits, merchants must rely on tools or systems. The personalized Recommendation System (RS) is commonly used to analyze customer’s purchase propensity, demand, and preference. RS can provide personalized products to satisfied customer; therefore, it can increase the percentage of returned customer. The combination of personalized recommendation and bundling strategy can increase not only the selling opportunity, but also trigger the customer’s purchasing desire. Moreover, it can increase the re-purchasing rate.
The past study of traditional product-bundling strategy tends to rely on merchant’s view. However, they do not think about customers’ demand and preference. In the long run, it may not increase customers’ purchasing desirability and it certainly can not meet the customer’s needs. Some researches [4] had proposed the combination of recommendation and bundling strategy, however, it did not provide the active recommendation for personalization. The timely recommendation research [6][32] had proposed in recent year, but it didn’t provide the recommendation of personalized product and consider the purchasing periodicity to satisfy the customers’ demand.
According to the above reasons, the bundling strategy to apply in RS effectively is designed. This research has proposed the mixed bundling strategy of recommendation to analyze customer’s periodical demand and preference. Customer’s purchase propensity and preference commodities were used for bundling. The goals of this research are (1) to improve the bundling strategy of products combination, (2) to find the periodical demands of product, (3) to provide a proper RS for customer, and (4) to use personal RS to improve the recommendation performance. The experiment results showed that the proposed bundling strategy did improve the sale rate.
|
|---|
