论文总字数：32710字

摘 要

信息技术飞速发展，互联网技术的运用日益成熟，这极大的促进了互联网行业的发展，“互联网 ”的提出促使众多传统行业向互联网行业快速转型。在此基础上，我国互联网平台发展起来，已有的互联网保险平台有大型保险公司自我经营的网站、第三方中介保险平台、电子商务保险平台和专业的互联网保险公司成立的平台。这些平台各具特色，不断探索新产品，较为有代表性的是特色保险。互联网保险产品有着碎片化、针对性强、标准化、定价低、交易频繁、合同简单易懂等特点。目前互联网保险平台还存在着产品数量少、推荐功能弱等缺点，针对此，本文从用户行为出发，探索适用于互联网保险产品的推荐算法。

首先，针对用户购买保险产品的行为进行了问卷调查，研究过程包括基于理论的问卷设计和问卷数据分析。问卷主要分为保险认知度与观点、互联网保险购买行为和个人基本信息三部分，通过数据分析，测出了用户个人因素对购买行为的影响程度，结果表明，用户年龄、收入对保险产品购买行为有显著影响，用户性别、文化程度、网购频率对购买行为影响不显著，但又正相关关系，具有一定的影响力。通过对产品因素对购买行为影响的分析，得出用户对产品类别、理赔信息、价格和服务水平重视度较高。第三章通过问卷数据分析，选取了影响互联网保险购买行为的用户因素与产品因素，将其应用到推荐算法中。

基于保险产品评论少、购买频率相对降低的特点，选取了协同过滤法与基于知识推荐方法组合的混合推荐法。首先对基于知识的推荐进行了改进，再将顾客浏览记录应用与基于协同过滤算法中。最后选取了切换混合推荐算法。很多用户第一次使用互联网保险产品，这类用户信息较少，可以利用基于知识的推荐算法进行推荐，对于信息较多的客户，采取协同过滤方法进行推荐。

本次研究基于用户行为分析，尝试性地提出来一个基于用户行为的互联网保险产品混合推荐算法，还未进行效果检验，后续会进行相关的探索。

关键字：互联网保险，电子商务，用户行为，个性化推荐

Ⅰ

Abstract

The use of Internet technology is becoming more and more mature. The new generation of information technology has greatly promoted the development of the Internet industry and has led to the rapid transformation of many traditional industries to the Internet industry. On this basis, China's Internet platform developed, the existing Internet insurance platforms include insurance companies self-site, third-party intermediary insurance platform, e-commerce insurance platform and platform set up by Internet insurance company. These platforms have their own characteristics and constantly explore new products. Internet insurance products are fragmented, targeted, standardized, low-price, and the contract is easy to understand. At present, internet insurance platforms still have some shortcomings, like limited product type, weak recommendation and so on. This article explore the recommended algorithm for Internet insurance products based on user behavior.

First, a questionnaire survey was conducted on the behavior of users purchasing insurance products. The research process included questionnaire-based questionnaire design and questionnaire data analysis. The questionnaire is divided into three aspects: insurance awareness and view, Internet insurance purchase behavior and personal basic information. Through the data analysis, the degree of influence of individual factors on purchasing behavior is measured. The results show that the user's age and income Behavior has a significant impact on the user gender, education level, online shopping frequency on the purchase behavior is not significant, but the positive relationship, has a certain influence. Through the analysis of the influence of product factors on the purchase behavior, it is concluded that the user's attention to the product category, claims information, price and service level is high. In the third chapter, through the analysis of questionnaire data, the user factors and product factors influencing the Internet insurance purchase behavior are selected and applied to the recommendation algorithm.

Based on the characteristics of fewer types of products and relatively low purchase frequency, a hybrid recommendation method based on collaborative filtering and knowledge recommendation method is selected. Firstly, the user-based collaborative filtering algorithm is improved, the collaborative filtering algorithm based on insurance products is improved, and then the application of knowledge-based recommendation method in insurance products is improved. Finally, the switching recommendation algorithm is selected. Many users use Internet insurance products for the first time. This kind of user information is less and can be recommended by using the knowledge-based recommendation algorithm. For the customers with more information, the cooperative filter is recommended.

This study is based on the user behavior analysis, trying to put forward a user behavior based on the Internet insurance product hybrid recommendation algorithm. But it has not yet an effect test, we will continue to explore.

Key Words: Internet insurance, E - commerce, user behavior, Personal recommendation

Ⅱ

目 录

摘要……………………………………………………………………………………………………………………………………………………………………Ⅰ

Abstract………………………………………………………………………………………………………………………………………………………………Ⅱ

第一章 绪论 1

1.1研究背景及意义 1

1.1.1研究背景 1

1.1.2研究意义 2

1.2国内外研究现状 2

1.2.1互联网保险行业 2

1.2.2推荐系统研究 2

1.3研究内容 3

1.4论文结构 4

第二章 理论基础 5

2.1行为保险学研究 5

2.2互联网保险平台与产品 5

2.2.1互联网保险平台 5

2.2.2互联网保险产品分类与特点 6

第三章 互联网保险用户行为分析 9

3.1互联网保险消费行为分析 9

3.2消费者互联网保险购买行为问卷设计 11

3.2.1问卷基本信息与设计原理 11

3.2.2问卷内容与结构 11

3.3影响用户购买行为的因素分析 12

3.3.1问卷数据分析 12

3.3.2影响用户行为的因素 14

第四章 基于行为的互联网保险产品推荐算法 16

4.1基于知识的推荐方法 16

4.1.1基于知识的推荐方法研究现状 16

4.1.2基于知识的推荐方法在互联网保险中的应用 16

4.2协同过滤推荐方法 17

4.2.1协同过滤方法研究现状 17

4.2.2改进的协同过滤方法 18

4.3混合推荐方法 18

第五章 案例应用 20

第六章 总结与展望 23

6.1 工作总结 23

6.2 工作展望 23

参考文献 24

附 录 26

致 谢 30

第一章 绪论

1.1研究背景及意义

1.1.1研究背景

随着社会经济的飞速发展，社会流动资金和家庭可支配收入不断增加，越来越多的人可以为各种风险做好防范储蓄，包括养老、教育、意外和可能的高额医疗费用[1]。与此同时，人们对于保险也有了新的认识，投保人的消费习惯也在发生改变，主动购买保险的人数飞速增长，2016年保险消费者信心指数结果显示：中国保险消费者信心指数为71.2，比中值50高出42.4%，相比2015年69.2的水平有所提升[2]，显示保险消费者信心进一步增强，保险购买量也大幅度增长，保险机构也逐渐增多，从大型保险公司到众多保险中介机构，到保险公司自创网站，到第三方保险销售平台，再到独立的互联网保险平台。

剩余内容已隐藏，请支付后下载全文，论文总字数：32710字