【作者】 余亮星；

【导师】 董爱华；

【作者基本信息】 东华大学 ， 控制理论与控制工程， 2011， 硕士

【摘要】 本文综合运用数据库管理、自动推荐等技术,从品牌服装零售店的实际需求出发,设计了基于RFID (Radio Frequency Identification)的服装零售店顾客服务系统,进行服装零售店管理,实现顾客服务提升的目的。该顾客服务系统的设计分为硬件和应用软件两部分。其中,硬件选取QR4401超高频固定式读写器及读写天线,选用LED显示器作为多媒体显示屏,电子标签则采用无源超高频电子标签。软件方面,采用SQL Server管理服装零售店系统的数据,并在Visual Studio平台上使用C#语言开发应用软件,实现标签维护、快速结算、智能试衣室、VIP顾客身份识别与自动推荐功能。文章首先从零售店的实际需求和功能出发,进行系统的硬件选型和选购,设计了底层数据库系统、系统的软件架构。然后,模拟系统维护人员工作场景,实现标签维护功能；模拟顾客购物结算场景,实现快速结算功能；模拟顾客试衣场景,实现智能试衣室功能；模拟VIP顾客光临场景,实现顾客身份识别与自动迎招。最后,提出一个复合推荐器,在VIP顾客光临时,触发生成个性化的推荐单,提供给顾客,供其在购物时参考。实验结果表明,本文所提出的方法是有效的,推荐单的结果与顾客的喜好基本是吻合的,贴切所有顾客的喜好程度可以达到73%,贴切单个顾客的喜好程度可以达到90%左右。本文所设计的基于RFID的服装零售店顾客服务系统来源于数字化纺织服装技术教育部工程中心项目,在服装零售店顾客服务系统上做了如下创新：在VIP会员卡里内嵌RFID电子标签,用于服装零售店的顾客身份识别与自动迎招。利用识别到的VIP顾客身份信息,创新性地在服装零售店中针对VIP顾客,应用了个性化的自动推荐功能。提出了一个复合推荐器,通过基于货品的协同过滤算法,获取与顾客历史购买相似度高的货品,加入初始推荐单；通过基于用户的协同过滤算法,获取喜好相近的其他会员顾客,提取其购买记录并加入初始推荐单；通过基于内容的过滤算法,获取初始推荐单中与顾客喜好最相似的货品,生成最终推荐结果,提供给顾客供其在购物时进行参考。更多还原

【Abstract】 By using database management technology and automatic recommendation technology, a Customer Service System based on Radio Frequency Identification Technology in Apparel Retailing Stores is designed according to actual demand of brand clothing stores, for the purpose of managing apparel retailing stores and improving customer service quality.The system hardware is made up of QR4401 UHF Fixed Reader, antenna, LED, computer and UHF passive tags. SQL Server is used for the data management of retailing clothing stores. Application software is developed in Visual Studio development platform by using C# language. Applications achieved are:tag maintenance, quick settlement, intelligent fitting rooms, VIP customer identification and automatic recommendation.First, according to the actual demand of clothing retail stores and its capability, the types of hardware devices are selected and purchased, database system and software architecture are designed. This paper then simulate system maintenance, and realize tag maintenance function; simulate shopping checkout, and realize quick settlement; simulate the scene of customer trying on clothes, and realize intelligent fitting room function; simulate loyalty customers’arrival, and realize customer identification and automatic greetings. Finally, a hybrid recommender is proposed in this paper, it is triggered to generate a personalized buying recommendation list when a loyalty customer who holds a RFID-circuit-embedded VIP card arrives at the store. The experimental results show that the selected recommender items are in consistence with the objective customers’interest, the relevance between the select recommender items and objective customers’ consumer preference can fit very well.The Customer Service System Project based on Radio Frequency Identification Technology in Apparel Retailing Stores is funded by Digital Textile and Apparel Technology Engineering Center, of the Minister of Education. Creative parts in the designed system are as follows:RFID-circuit embedded VIP cards are engaged to identify VIP customers for their arrivals, hence automatic greeting could be offered to these customers.Automatic personalized recommendation is creatively applied for VIP customers, by using the indentified identity information of VIP customers. A hybrid recommender is proposed, which could generate personalized recommender list to the VIP customers. Particularly, using the item-based filter algorithm, the products that have high similarity with those VIP members have bought recently are selected into the initial recommender list; using the user-based filter algorithm, the products that other VIP members who have the same tastes with the objective customer have chosen recently are added into the initial recommender. Finally, contented-based recommender algorithm is engaged to generate the final recommender list.更多还原