在线推荐系统消费者采纳意向的影响机理研究

发布时间：2018-08-03 10:54

【摘要】：近年来,在电商网站的带动下,推荐系统不断发展。目前关于推荐系统的研究多集中于推荐系统算法设计优化方面,从消费者角度研究推荐系统的相关文献较少,本研究从消费者角度出发研究推荐系统对消费者采纳意向的影响机理,将有助于完善推荐系统设计,提高推荐服务质量。本文基于消费者视角,结合技术接受模型和信息系统成功模型,以推荐系统特征为自变量,构建推荐系统对消费者采纳意向影响模型,分别以手机和图书为搜寻品和体验品的代表,通过问卷调查法收集有效问卷386份,并借助smart PLS软件对研究模型进行结构方程检验分析,最后对完善推荐系统提出建议。主要结论如下:(1)推荐系统可以从信息质量、系统质量和交互质量三个方面进行评价,信息质量可以从推荐准确性、推荐多样性、推荐新颖性、推荐关联性等方面进行衡量,系统质量可以从界面设计和推荐解释等方面进行衡量,交互质量可以从系统交互质量和用户间交互质量等方面进行衡量;(2)对于搜寻品来说,推荐准确性、系统交互质量、推荐多样性、推荐关联性、界面设计对采纳意向既有间接影响,也有直接影响。而对于体验品来讲,推荐准确性、推荐解释、界面设计、系统交互质量、用户间交互质量对采纳意向有间接影响,而且也有直接影响;(3)对于搜寻品来说,推荐系统特征对采纳意向的影响程度由大到小依次是系统交互质量、推荐准确性、推荐多样性、界面设计、推荐关联性。对于体验品,推荐系统特征对采纳意向的影响程度由大到小依次是用户间交互质量、界面设计、推荐解释、推荐准确性、系统交互质量;(4)推荐新颖性对采纳意向的影响作用在两类产品中都不显著;(5)建议商家加深对推荐系统角色的认知,根据不同的产品类别设计推荐系统,并拓展推荐系统的交互功能。
[Abstract]:In recent years, under the impetus of ecommerce website, recommendation system develops continuously. At present, most of the researches on recommendation system are focused on the optimization of the algorithm design of the recommendation system, and there are few related documents to study the recommendation system from the consumer's point of view. This study studies the influence mechanism of recommendation system on consumers' intention from the perspective of consumers, which will help to perfect the design of recommendation system and improve the quality of recommendation service. In this paper, based on the consumer perspective, combining the technology acceptance model and the information system success model, taking the characteristics of the recommendation system as the independent variable, this paper constructs the model of the impact of the recommendation system on the consumer's intention to adopt. Taking mobile phone and books as the representatives of search and experience, 386 valid questionnaires were collected by questionnaire, and the structural equation of the research model was tested and analyzed by smart PLS software. Finally, some suggestions were put forward to improve the recommendation system. The main conclusions are as follows: (1) recommendation system can be evaluated from three aspects: information quality, system quality and interaction quality. System quality can be measured from interface design and recommendation interpretation, interaction quality can be measured from system interaction quality and user interaction quality. (2) for search products, recommended accuracy, system interaction quality, etc. Recommendation diversity, recommendation relevance and interface design have both indirect and direct effects on the adoption intention. For experience products, recommendation accuracy, recommended interpretation, interface design, system interaction quality, and user interaction quality have indirect effects on the adoption intention, and also have a direct impact on the adoption intention; (3) for search products, The influence of the features of the recommendation system on the intention of adoption is followed by the quality of system interaction, the accuracy of recommendation, the diversity of recommendation, the design of interface and the relevance of recommendation. For the experience, the influence of the features of the recommendation system on the intention of adoption is in order of the interaction quality, interface design, recommendation interpretation, recommendation accuracy between the users. System interaction quality; (4) the effect of recommendation novelty on adoption intention is not significant in both categories of products; (5) it is suggested that merchants deepen their understanding of the role of recommendation system and design the recommendation system according to different product categories. And expand the interactive function of recommendation system.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.3

【参考文献】