BAYESIAN ANALYSIS OF CROSS-CATEGORY ATTRIBUTE PREFERENCES: PERSONALIZED PRODUCT RECOMMENDATIONS

跨品類產品屬性偏好之貝氏統計分析：個人化產品推薦

劉秀雯(Hsiu-Wen Liu) ； 任立中(Lichung Jen) ； 邵功新(Kung-Hsin Shao) ； 林育理(Yu-Li Lin)

《品質學報》 24卷5期 (2017/10) Pp. 360-371

https://doi.org/10.6220/joq.2017.24(5).05

摘要

When choosing which product to recommend to a target customer, firms often rely upon content-based or collaborative filters that either do not account for the heterogeneity of their target market or do not consider the trade-offs that consumers are willing to make for different product options. In this research, we develop a framework for investigating individual consumer preferences. This framework incorporates two steps. First, the Bayesian Variable Selection method is employed in order to select important variables. Second, a Hierarchical Bayes Probit model is developed in order to reflect the heterogeneity of individual preferences. Our empirical results demonstrate that the proposed method performs well in terms of discovering individual preferences toward cross-category common attributes at the abstract level. These findings provide important insights for retailers currently looking for ways to differentiate themselves using personalized product recommendations.

關鍵字

Bayesian variable selection ； product recommendation ； hierarchical Bayes modeling

並列摘要

當企業想針對特定顧客進行產品推薦活動時，他們通常仰賴以內容為基礎或協同過濾法等演算法來分析，然而這些分析方法通常沒有考慮消費者偏好的異質性，也無法反應顧客在面對選擇時的取捨行為，故無法探討個別顧客偏好或選擇行為的差異。在本研究中，作者提出了一套分析架構來估計顧客個人化的偏好結構，並運用於個人化的新產品推薦情境。此分析架構包含兩個重要的步驟。首先，我們透過貝氏變數選擇方法來選擇重要的產品屬性變數。其次，我們提出了層級貝氏統計模式來估計個人化的偏好結構參數。本研究的實證結果顯示所提出的方法對於跨品類個人化偏好結構的估計與推薦成效是良好的。本研究對於想提升個人化推薦品質與績效的零售商而言具有豐富的管理策略意涵。

並列關鍵字

貝氏變數選擇分析 ； 產品推薦 ； 層級貝氏模式

參考文獻

Adomavicius, G. and Tuzhilin, A., 2005, Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions, IEEE Transactions on Knowledge and Data Engineering, 17(6), 734-749.

Aichner, T., 2014, Country-of-origin marketing: a list of typical strategies with examples, Journal of Brand Management, 21(1), 81-93.

Allenby, G. M., Bakken, D. M., and Rossi, P. E., 2004, The HB revolution: how Bayesian methods have changed the face of marketing research, Marketing Research, 16(2), 20-25.

Allenby, G. M. and Ginter, J. L., 1995, Using extremes to design products and segment markets, Journal of Marketing Research, 32(4), 392-403.

Ansari, A., Essegaier, S., and Kohli, R., 2000, Internet recommendation systems, Journal of Marketing Research, 37(3), 363-375.