在竞争激烈的网上零售活动中,为了提升消费者的购物体验、培养顾客忠诚,越来越多的电子商务企业开始为顾客提供个性化商品推荐。目前个性化推荐相关文献主要研究如何改进推荐算法、提升推荐质量,而关于个性化推荐与消费者行为之间关系的研究相对较少。这导致学者和电子商务从业人员过于关注推荐算法本身,而忽略了推荐对消费者和电子商务企业的影响。为了帮助学界和业界更好地认识和理解个性化推荐及其研究脉络,本文从个性化推荐的定义和分类、推荐算法和模型、个性化推荐与消费者行为之间的关系这三个方面对现有文献进行了系统的梳理。与现有的相关文献综述不同,本文侧重于探讨个性化推荐与消费者行为之间的关系,主要包括消费者对个性化推荐的评价及其影响因素以及个性化推荐对消费者网上购买决策过程、决策结果的影响。最后,本文提出了将来的一些研究方向,供营销和信息科学领域的学者进一步研究和探讨。
In fiercely competitive online retail activities, to improve customer shopping experience and cultivate consumer loyalty, more and more e-commerce websites now begin to provide personalized recommendations for their customers. The earliest studies regarding personalized recommendations focus on the improvement of algorithms or models that improve recommendation quality. However, relatively few studies examine the relationship between personalized recommendations and customer online shopping behavior. It makes scholars and e-commerce practioners place more emphasis on recommendation algorithm itself but neglect the effects of recommendations on consumers and e-commerce websites. To help people better understand personalized recommendations and related research context, this paper comprehensively reviews related literature in three research streams. The first research stream is the definition and classification of personalized recommendations. The second research stream includes investigations and explorations on the recommendation algorithms and statistical models that improve the recommendation quality. The third research stream includes literature on the relationship between personalized recommendations and customer shopping behavior. Different from current literature review, it places emphasis on the relationship between personalized recommendations and consumer shopping behavior, including evaluation of personalized recommendations by consumers and influencing factors and the effects of personalized recommendations on the process and results of consumer online shopping decisions. Finally it provides marketing and information science scholars with some directions in future research.