應用協同過濾法設計信用卡點數兌換之推薦系統

Using Collaborative Filtering to Design a Recommendation System based on Credit Card Points

林文新(Wen-Hsin Lin)

指導教授 ： 黃有評

大同大學/電機資訊學院/資訊工程學系所/碩士(2008年)

摘要

在現今的社會中，由於信用卡攜帶方便且刷卡付費可延遲繳款時間，使得眾多的場合，舉凡餐廳、購物中心、旅行的機票住宿等，幾乎皆可刷卡付費，再加上刷卡付費可累積紅利點數兌換免費商品，人手數張信用卡已成為社會上的一種趨勢。然而，大多數的消費者在刷卡付費後，往往忘記累積的紅利點數是可兌換免費商品的，而浪費了兌換商品的機會。此外當消費者面對眾多可兌換的商品時，往往對每件商品都很喜歡，猶豫不決到底兌換哪樣商品好，因此我們提出了一個具有推薦機制的兌換系統，讓消費者不用煩惱這問題，由系統自動告知消費者適合兌換的商品，且能隨時隨地皆能輕鬆的察看及兌換商品。
不同於傳統兌換商品的系統，我們將網頁伺服器及資料庫和網路作結合，並加入協同過濾推薦機制及個人化的技術，透過分析每位消費者察看及兌換商品的紀錄，找出其個人的偏好，並為其建立一個專屬的個人化推薦機制。讓消費者不論身在何處只要有網路及電腦或手持裝置，便可透過網頁瀏覽器登入此系統網站，進而清楚地知道以目前自己所擁有的點數可兌換哪些最適合的商品及再累積多少點數後可兌換更高級的商品。在此基礎下，除了證明我們系統的推薦準確性、整體可行性外，也希望透過網路的普及性及系統內簡易的操作介面，使得此系統更加的大眾化，進而可真正的運用在市面上銀行的網站中，增加兌換商品的便利性。

關鍵字

協同過濾法 ； 信用卡點數 ； 推薦機制

並列摘要

In today's society, due to the credit card is easy to carry and buying products by using credit card can delay the time of payment, it makes people consume by credit cards in many occasions, such as restaurants, shopping mall, the flight tickets and traveling. In addition, consuming by credit card can accumulate the points to exchange the free products. However, most consumers may forget that the accumulated credit points can be used to exchange the free products. When consumers facing the multiple choices of free products, they often like all products and do not know which products to exchange are better. So, a recommendation system is proposed. The system will tell the consumer which products he/she can exchange appropriately and the consumer can browse and exchange the products easily at any time.
Different from the traditional system within the bank web, the proposed system combines web server, database and Internet technology with collaborative filtering and recommendation model. It finds the customers’ preference by analyzing their browsing and exchanging products records and then builds a personal recommendation system. The consumer can use PC or hand-held devices to browse the web and check what items he/she can exchange under the available credit points and what items he/she may exchange in the future if more points are accumulated.
On the basis of this concept, the experiment results have verified the feasibility of the overall recommendation system. We hope that our design becomes more popular through the popularity of the Internet. And then, this system can be really used in the bank's website in the market, and increase the convenience of exchanging products.

並列關鍵字

collaborative filtering ； credit card point ； recommendation

參考文獻

[2] G. Linden, B. Smith and J. York, “Amazon.com recommendations: item-to-item collaborative filtering,” IEEE Internet Computing, vol. 7, issue 1, pp.76-80, Jan.-Feb. 2003.

[6] H.-H. Sung, “Digital content recommender on the Internet,” IEEE Intelligent Systems, vol. 21, no. 2, pp.70-77, Mar.-Apr. 2006.

[7] B. Zhou, S.-C. Hui and K. Chang, “Enhancing mobile web access using intelligent recommendations,” IEEE Intelligent Systems, vol. 21, no. 1, pp.28-34, Jan.-Feb. 2006.

[13] R.J. Mooney and L. Roy, “Content-based book recommending using learning for text categorization,” Proc. of the 5th ACM Conf. on Digital Libraries, San Antonio, TX, USA, pp.195-204, August 2000.

[17] E. Ohbuchi, H. Hanaizumi and L.-A. Hock, “Barcode readers using the camera device in mobile phones,” Proc. of IEEE Int. Conf. on CYBERWORLDS, Tokyo, Japan, pp.260-265, Nov. 2004.