以虛擬經驗為基礎之個人化產品推薦系統建置

A Study on Constructing a Virtual Experience Based Personalized Product Recommendation System

王婷儀(Ting-Yi Wang)

指導教授 ： 耿慶瑞

國立臺北科技大學/管理學院/商業自動化與管理研究所/碩士(2007年)

https://doi.org/10.6841/NTUT.2007.00313

摘要

虛擬經驗就是消費者透過電子媒體模擬的方式在網際網路上感受到的直接產品經驗。網際網路提供了一個平臺，讓消費者在搜尋、比較及存取產品資訊時更加便利，且能獲得比實體商店更深一層的產品資訊，因此網站將可以用經驗形式推產品給消費者。但目前相關推薦系統的研究，大多著重個人化技巧或資料探勘;或是只有將產品或品牌名稱的推薦，未考慮以經驗形式的推薦系統，而虛擬經驗，可以根據參與程度及沉浸程度兩個構面，將分為教育經驗、娛樂經驗、美感經驗及逃避現實經驗四種類型(Pine and Gilmore, 1999)。事實上，以經驗形式將推薦系統呈現給消費者會更具有說服力(Hoch and Deighton, 1989)，因此以虛擬經驗為基礎的推薦就有其重要性。故本研究認為推薦系統的研究，應該要跳脫過去資訊過濾與鑽研演算法求取最精確的顧客命中為目的，以本研究嘗試將「協同過濾」結合「以規則為主的過濾」機制應用在網站商品個人化推薦上，這種技術是藉由結合一些志同道合的使用者，利用其偏好資料及條件篩選，以推薦使用者喜愛產品資訊。
本研究結果顯示結合「個人化虛擬經驗」及「個人化產品推薦」的虛擬經驗產品推薦系統跟沒有推薦系統比較後，發現消費者的時間價值、滿意度及購買意願皆得到提升。

關鍵字

虛擬經驗 ； 推薦系統 ； 協同過濾機制

並列摘要

It is the virtual experience that the direct products experience consumers acquire from Internet through electronic media simulation. What Internet provides is a platform which makes the consumers more convenient for searching, comparing, and accessing products information. Besides, consumers would obtain deeper products information from Internet than actual shops.

Experience form could be used to recommend products to consumers through Internet. Until now, the researches about recommendation system emphasized more on personalized skill and data mining or only on products name or brand title recommendation without deliberating about recommendation system using experience form. According to the participative degree and enchanted grade, we distinguish virtual experience into four types, such as educational experience, recreational experience, aesthetic experience, and experience of shrinking fact (Pine and Gilmore, 1999). In fact, it is of persuasion displaying the recommendation system to consumers by experience form (Hoch and Deighton, 1989), it is pretty important for recommendation mechanism based on virtual experience.

Not should recommendation systems persist in focusing on precise customer selection by filtering information and delving algorithm. Therefore, we attempt to adopt a personalized recommendation mechanism for website merchandise by combining collaborative filtering with rule-based filtering. The technique we introduce is accomplished by certain users who are in the same camp provide the information of fond products for consumers by taking the preference data and filtering condition. This research finally reveal higher value on consumers’ time worth, satisfied degree, and shopping pleasure in comparison between the virtual experience personalized products recommendation system, which integrates personalized virtual experience with personalized products recommendation mechanism, and several systems without recommendation mechanism sustenance.

並列關鍵字

Virtual Experience ； Recommendation System ； Collaborative filtering

參考文獻

[4] 周晏頤，經驗學習策略對虛擬經驗類型之廣告效果的影響，碩士論文，台北科技大學商管所，2007年。

[5] 劉智華，網站體驗與上站忠誠度之關係研究－以資訊提供型網站為例，碩士論文，中原大學資訊管理研究所，2001 年。

[2] 楊郁儀，觸覺需求與產品涉入對消費者經驗順序組合之學習效果與認知風險影響，碩士論文，台北科技大學商管所，2005年。

[3] 姚昭如，美學鑑賞程度與獨特性傾向對消費者經驗順序組合之學習效果影響，碩士論文，台北科技大學商管所，2005年。

[1]Alba J., John L., Barton W., Chris J., Richard L., Alan S., & Stacy W. (1997). Interactive Home Shopping: Consumer, Retailer, and Manufacturer Incentives to Participate in Electronic Marketplaces. Journal of Marketing, 61(3), 38-53.