论文总字数：26239字

摘 要

近年来，随着互联网的普及，我国电子商务发展突飞猛进，各大电子商务平台销量年年上涨，各种促销活动层出不穷。与此同时，也带来了一些问题与挑战，在信息不对称的虚拟交易平台上购买商品，如何确认卖家的可信度以寻找值得信赖的店铺购物是众多问题中非常重要的一个。不仅买家关心此问题，电子商务平台也同样关注这个问题，各电商平台都有自己的电子商务信用评价系统，对卖家的信用进行评级，以供买家参考和作为平台考核活动资格等的依据。

本文以淘宝平台C2C店铺为例，分析其信用评价模式的不足之处，通过对信用指标数据进行聚类分析，归纳总结，选取指标构建电子商务平台店铺信用评价模型，该模型涵盖了保障性维度的店铺开店时长、金牌卖家认证、保险公司承保、保证金指标，口碑维度的好评率、描述相符评分、服务态度评分、物流服务评分指标，以及流量维度的销售量指标，根据店铺在不同维度的表现情况分为高低档次，再由三个维度的不同组合将店铺分为6个等级，用以衡量店铺的信用水平。

该模型整合了各信用指标，建立多维信用指标体系，全面系统考评店铺信用，降低单一指标对信用水平的决定性影响，在一定程度上解决了淘宝网现有评价模式的一些问题。

关键词：信用评价，电子商务，聚类分析，C2C

ABSTRACT

In recent years, with the popularization of the Internet, e-commerce in China has developed rapidly, and sales of major e-commerce platforms have risen year after year, and various promotional activities have emerged one after another. At the same time, it also brings some problems and challenges. Buying goods on an asymmetric virtual trading platform, how to confirm the credibility of the seller to find trustworthy store shopping is a very important one among these issues. Not only does the buyer care about this issue, the e-commerce platforms also pay attention to this issue. Each e-commerce platform has its own e-commerce credit evaluation system, rating the seller's credit for the buyer's reference and as a basis for platform assessment activities.

Taking Taobao platform C2C store as an example, this paper analyzes the shortcomings of its credit evaluation model. Through clustering analysis of credit index data, summarizes and selects indicators to build a store credit evaluation model for e-commerce platform. The model covers the length of shop opening, gold medal seller certification, insurance company underwriting, guarantee deposit in security dimension, praise rate, description conformity score, service attitude score, logistics service score in word-of-mouth dimension, and sales index in traffic dimension. According to the performance of the shop in different dimensions, it is divided into high and low grades, and then the shops are divided into 6 levels by different combinations of three dimensions to measure the store's credit level.

The model integrates various credit indicators, establishes a multi-dimensional credit indicator system, and systematically evaluates store credit. Reducing the single index's decisive influence on the credit level solves the defect of Taobao's existing evaluation model to some extent.

KEY WORDS: Credit Evaluation, Electronic Commerce, Cluster Analysis, C2C

目 录

摘要 Ⅰ

Abstract Ⅱ

第一章 绪论 1

1.1研究背景 1

1.2选题目的和意义 4

1.3研究思路与方法 5

1.4创新点 5

第二章 相关理论及文献综述 6

2.1信用评价 6

2.2文献综述 7

2.2.1国外文献综述 7

2.2.2国内文献综述 9

第三章 聚类分析 13

3.1数据选取 13

3.1.1店铺指标 14

3.1.2商品指标 14

3.1.3服务指标 15

3.2聚类实现 15

3.3聚类结果 17

第四章 基于聚类结果的信用模型 19

4.1类别分析 19

4.1.1第一类 19

4.1.2第二类 19

4.1.3第三类 19

4.1.4第四类 19

4.1.5第五类 20

4.1.6总结 20

4.2信用模型 20

4.2.1分类规则 20

4.2.2信用等级 21

4.2.3模型优势 22

第五章 总结与展望 24

5.1总结 24

5.2展望 24

参考文献（References） 26

附录 30

致谢 32

第一章 绪论

1.1研究背景

二十世纪九十年代以来，随着互联网技术的发展和普及，互联网用户数量激增，网上购物逐渐流行，尤其在中国，阿里巴巴等电商巨头的崛起，使得网上购物成为人们的一大生活习惯，甚至成为一种中国特色。当前我国电子商务行业发展已进入成熟期，增速减缓，市场规模趋向稳定，根据国家统计局电子商务交易平台的调查显示，电子商务2018年全国交易额为31.63万亿元，比2017年增长了8.5%。其中商品、服务类电商交易额30.61万亿元，增长14.5%；合约类电商交易额1.02万亿元，下降51.3%。

剩余内容已隐藏，请支付后下载全文，论文总字数：26239字