| Summary: | 碩士 === 國立雲林科技大學 === 資訊管理系碩士班 === 89 === Marketing evolves increasingly from production era into relationship marketing nowadays. Enterprises emphasis on customer-oriented and uses allover customer relationship management strategies to retain the profitable customers, then establish the close relationship between customers, finally achieve the goal of relationship marketing. Before implementing the customer relationship management, enterprises have to understand customers’ needs and patterns and design proper relationship strategies base on individual characteristics. With the popularity and the maturity of databases, most organizations store lots of data about consumers but hardly discover the deep, unknown meanings, further transform them into valuable information or knowledge, and use these data to support the integrated decision-making about sales, marketing and customer service. We propose one analysis model, discovery the potential information via the data mining techniques, organizations could use the information to support marketing decision-making and design customer relationship management strategies. In the analysis processes, the data are preprocessed first, then cluster the data depending on members’ attributes by clustering technique and extracts the demographic and consumption patterns for each cluster, finally presents the associations between patterns via association rules. Besides the patterns, we analyze the product categories, amount and frequency of members’ transactions, attempt to discovery the special transaction-trigger events instead of changing information. This research verifies the analysis model by developing a prototype and using the real transaction records.
|
|---|
