基于层次分析法的服装推荐专家系统

发布时间：2018-05-03 02:08

本文选题：服装智能推荐 + 专家系统　；参考：《东华大学》2016年硕士论文

【摘要】：随着我国经济的发展,服装产业尤其是网上购衣也取得了迅速发展。线上的B2C模式C2C模式现在更为年轻的消费群体所青睐。现在越来越多的顾客选择在一些购物网站如淘宝,京东等购物,网络购物俨然已经成为了一种大众购物方式。网上有许多服装店铺同时进行线上线下同步销售。这种网络销售方式,无需实体店铺,省去了店铺租赁费用,同时,由于这种网上店铺不局限于地理范围,扩大了客源,因此取得了很快的发展。但就网上购买服装而言,还存在一个问题:消费者在这些网络店铺并不能接触衣服,仅能通过商家在网页上的描述了解服装的款式、材质和颜色,但没有经过试穿,仅仅凭借图片和尺码推荐表很难选到合适的款式和大小。这一问题减少了一定的客流量,制约着服装电子商务的进一步发展。基于以上情况,本课题基于专家系统和层次分析法进行服装智能推荐系统的设计,包括服装款式推荐和服装尺码推荐。本文提出了一种添加索引表的黑板结构,并将其应用在服装款式推荐专家系统中,以产生式规则建立服装搭配专家规则库,通过人机交互获取顾客的个人信息,从服装专家的角度,生成个性化服装款式推荐单,供顾客选择。此外,将层次分析法应用于服装尺码推荐,这一部分着重建立了服装型号递阶层次结构,选择身高、胸围、臂长及肩宽控制参数作为准则层,具体服装供选择的尺码作为方案层,通过对判断矩阵的计算以及层次排序给出最终尺码选择。本文本系统软件开发采用C/S架构,利用Mysql和Java来开发。后台数据库基于mysql,存放顾客和服装信息数据,并用作规则库、事实库、动态数据库来存放服装推荐规则、事实及推理的中间结果等;前台则基于Java语言开发应用软件,实现服装智能推荐系统。本系统节约了网购顾客的购物时间,简化了购物流程,使顾客在购物消费中体验到针对自己的智能化、个性化服务,促进了服装零售业智能化发展,一定程度上消除其对网购失败的顾虑,提升了顾客体验,对服装电子商务的发展具有一定的促进作用。
[Abstract]:With the economic development of our country, clothing industry, especially online clothing, has also made rapid development. The online B-C-C model is now favoured by younger consumer groups. Now more and more customers choose to shop in some shopping websites such as Taobao JingDong and so on. Online shopping has become a popular shopping method. There are many online clothing stores at the same time online and offline synchronous sales. This kind of network sale way, does not need the physical shop, saves the shop rental expense, at the same time, because this kind of online shop is not limited to the geographical scope, has expanded the passenger source, therefore has obtained the fast development. But when it comes to buying clothes online, there's a problem: consumers don't have access to clothing in these online stores, they can only learn about the styles, materials and colors of clothing through their online descriptions, but they don't try it on. It is difficult to choose the right design and size based on the picture and size table alone. This problem has reduced the certain passenger flow, restricted the clothing electronic commerce the further development. Based on the above situation, this paper designs the clothing intelligent recommendation system based on expert system and Analytic hierarchy process (AHP), including fashion style recommendation and garment size recommendation. In this paper, a blackboard structure of adding index table is put forward, and it is applied to the fashion style recommendation expert system. The production rule is used to establish the dress collocation expert rule base, and the personal information of the customer is obtained through man-machine interaction. From the perspective of clothing experts, to generate personalized clothing style recommendations for customers to choose. In addition, the Analytic hierarchy process (AHP) is applied to recommend the size of clothing. In this part, the hierarchical structure of clothing model is established, and the control parameters of height, chest circumference, arm length and shoulder width are selected as the criterion layer. The size of the specific garment is chosen as the scheme layer, and the final size selection is given through the calculation of the judgment matrix and the ranking of the hierarchy. This paper uses C / S architecture, Mysql and Java to develop the system software. The backstage database is based on mysql, stores customer and clothing information data, and is used as rule base, fact base and dynamic database to store clothing recommendation rules, facts and intermediate results of reasoning, etc. The intelligent clothing recommendation system is realized. The system saves the shopping time of online shopping customers, simplifies the shopping process, makes customers experience their own intelligent, personalized services in shopping consumption, and promotes the intelligent development of clothing retailing. To a certain extent, it removes its concerns about the failure of online shopping, improves the customer experience, and promotes the development of apparel e-commerce to a certain extent.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.3

【相似文献】