隨著網際網路與資訊科技的發展,行銷手法因為環境的改變及顧客關係的管理等因素而產生變化。近來資料庫行銷的興起,讓許多企業公司紛紛建構資料庫,資料庫內容包含顧客的基本資料、交易時間以及交易的項目…等,但更重要是後續的資料分析動作以提供決策之參考。適用之行銷輔助模式不僅能提供行銷決策參考、提升行銷效益、降低行銷成本,並且能達到一對一之最高行銷策略使顧客滿意,以獲得更大利潤。本研究以案例式推理方法與自組織映射圖網路兩種方法,藉由比對現有資料庫內顧客的基本資料,找出與新顧客之相似顧客,利用相似顧客之購買模式來推薦新顧客適合的產品。比較兩種方法所找到與新顧客相似之顧客群對於新顧客之消費模式的預測正確率,結果為案例式推理方法之平均預測正確率(約53%)優於自組織映射圖網路方法之平均預測正確率(約33%),約大於20%。因此,案例式推理方法適合於本化妝品業之研究案例在建構行銷輔助模式之應用。
With the developing of Internet and information technology, the promotion method, marketing environment and the relationship with customers are changing. With the rising of database marketing, many companies begin to construct their database. Usually the database includes the basic data of the customers, exchange time and exchange items. One of the most important tasks is the data analysis and that a suitable promotion model not only can support the decision making, promote marketing effect, reduce cost but also attain the high customer satisfaction to earn more profit. In this research, we proposed a method that uses the customers with similar attributes to the new customers from the real data in cosmetics in order to promote products and services based on what they purchased to new customers. Two methods were applied, i.e. the Case Based Reasoning and the Self Organizing Map neural network. They were applied to forecast the demands of the new customers. The comparison of the two methods was done by using real data. The results show that the forecasting accuracy of the Case Based Reasoning (53%) is 20% more than that of Self Organizing Map (33%). The Case Based Reasoning method can derive better performance than the Self Organizing Map method in this cosmetics case for making the auxiliary promotion model.