產品推薦是線上購物中常見的功能,推薦之相關產品通常在畫面中列於主要產品下方,但畫面上也有較不相關的廣告產品。由文獻已知消費者對相關性高的產品感受較佳,對較不相關的產品感受較差,進而影響了消費者的購買態度。因此,產品相關性與消費者感受之間的關係十分密切且重要,然而目前仍未有足夠的文獻探討兩者的關係。所以本研究主題為探討消費者的購物經驗如何隨著推薦產品相關程度的變化而改變。方法第一階段為確認產品間的相關程度,以20位受測者分類100項產品的結果建構出群集樹狀圖,再依據群集結構分成最高、高、中、低與最低等五階產品相關程度,即為本研究之自變數,第二階段再以此五階相關程度分別製作五種推薦產品畫面,每種畫面皆有一項主要產品和四項推薦產品,讓40位受測者在無時間限制下瀏覽這五種畫面,並以購物經驗問卷衡量他們瀏覽推薦產品時對於推薦資訊品質、購物攸關性及感知有用性的感受,即為本研究之應變數。本研究結果為,不論中到低相關或高到中相關的購物經驗降幅,都顯著大於最高到高相關以及低到最低相關之降幅。此趨勢應可以邏輯斯(logistic)方程式解釋,因此利用曲線估計模型得出產品相關性與購物經驗之迴歸方程式,此迴歸方程式可讓推薦系統設計者了解產品相關性與線上購物經驗之關係,將能有助於設計策略之訂定。
Recommended products are commonly displayed around target products on shopping websites. Most of the recommended products relate to the target products, but some advertised products may not be so relevant. Previous research has shown that the relevance of recommended products to target products affects customers’ shopping experience and attitude. The higher the relevance is, the more positive the experience and attitude will be. Therefore, the relationship between relevance of recommended products and shopping experience is important. However, insufficient research has been done to explore this relationship. The purpose of this paper is to investigate how customer’s shopping experience changes to recommended products according to different degrees of relevance to target products. To obtain the relevance of products, twenty participants were asked to sort 100 products so that a dendrogram about product relevance was constructed. Based on the dendrogram, 5 levels of relevance were identified. They were the highest relevance, high relevance, medium relevance, low relevance and the lowest relevance, and served as the 5 levels of the independent variable. Five shopping webpages were designed according to these 5 levels of relevance. For each webpage, a target product displayed on the center of the page and 4 recommended products horizontally listed below it. Another40 participants were asked to browse all the five webpages in a random order and to complete a shopping experience questionnaire for measuring participants’ shopping experience of recommendation information quality, shopping relevance and perceived usefulness. Results indicated that the declines of shopping experience from medium to low relevance and from high to medium relevance were significantly sharper than the declines from highest to high relevance and from low to the lowest relevance. A logistic function should be able to describe the relationship between the shopping experience and product relevance. Thus, through the curve estimation modeling, a logistic regression function was derived. It is expected that designers of recommendation systems can apply this function to their design decisions.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。