本文選題：數(shù)據(jù)挖掘　切入點：Web日志挖掘　出處：《天津大學(xué)》2014年碩士論文

【摘要】：如今,隨著Web技術(shù)的不斷發(fā)展,電子商務(wù)站點風(fēng)起云涌,其已經(jīng)成為商業(yè)領(lǐng)域內(nèi)的新模式。電子商務(wù)網(wǎng)站,就技術(shù)角度來說,開發(fā)部署一個電子商務(wù)站點并不是什么難事,關(guān)鍵問題是如何讓自己的電子商務(wù)站點能夠獲得更高的盈利,而Web挖掘技術(shù)正是針對這一問題,并能夠有效的對其進(jìn)行解決。Web日志挖掘通過對用戶訪問Web服務(wù)器時留在服務(wù)器上的Web日志進(jìn)行分析,使用數(shù)據(jù)挖掘技術(shù),發(fā)現(xiàn)隱藏在Web日志中的規(guī)律性知識,就可以得到用戶的訪問模式以及興趣偏好。通過Web日志挖掘可以為構(gòu)建個性化服務(wù)提供依據(jù)以及為電子商務(wù)的開發(fā)和為管理人員提供大量有利于對Web站點改進(jìn)、帶來經(jīng)濟(jì)效益的信息。本文給出了Web日志挖掘技術(shù)的應(yīng)用框架。然后,以該框架以及Web日志挖掘的一般框架為基礎(chǔ),以理解客戶為目標(biāo)設(shè)計了電子商務(wù)Web日志挖掘的一種簡易框架圖,并把整個挖掘工作分為三個部分,即數(shù)據(jù)清理、模式挖掘以及模式分析,便于對實際應(yīng)用的操作。由于Web日志挖掘的流程相對繁瑣,特別是數(shù)據(jù)預(yù)處理工作。因此,本文以簡化圖為指導(dǎo),采取了兩種新的XML應(yīng)用,即XGMML和LOGML,可以使預(yù)處理工作大大地簡單化；以此為基礎(chǔ),對商品的客戶購買模式進(jìn)行挖掘,既包括客戶的個體訪問模式,也包括群體訪問模式。這不僅有利于實現(xiàn)“一對一”的個性化服務(wù)要求,同時,也運(yùn)用了分類技術(shù)對客戶群體進(jìn)行了劃分,提升了商家的營銷準(zhǔn)確性。最后,以上述的訪問模式的挖掘技術(shù)為前提,對Web挖掘模式在電子商務(wù)環(huán)境下的客戶分類支持框架進(jìn)行了詳細(xì)地分析,以及對電子商務(wù)環(huán)境下的客戶管理問題進(jìn)行了重點討論,以此來正確地制定企業(yè)營銷策略,幫助企業(yè)吸引客戶,提高盈利。
[Abstract]:Today, with the continuous development of Web technology, e-commerce sites have become a new model in the business field. In terms of technology, it is not difficult to develop and deploy an e-commerce site. The key question is how to make your e-commerce site more profitable, and Web mining technology is aimed at this problem. By analyzing the Web logs left on the server when the user visits the Web server, and using the data mining technology, the regular knowledge hidden in the Web log can be found. Through Web logging mining can provide a basis for building personalized services, e-commerce development, and a large number of Web site improvements for administrators. This paper gives the application framework of Web log mining technology. Then, based on this framework and the general framework of Web log mining, A simple framework diagram of Web log mining in electronic commerce is designed to understand customers, and the whole mining work is divided into three parts: data cleaning, pattern mining and pattern analysis. The flow of Web log mining is relatively cumbersome, especially the data preprocessing work. Therefore, two new XML applications are adopted under the guidance of simplified graph. In other words, XGMML and LOGML can greatly simplify the preprocessing work. On this basis, we can mine the customer purchase patterns of commodities, including the individual access patterns of the customers. This not only helps to realize the "one-to-one" personalized service requirements, but also uses the classification technology to divide the customer groups, which improves the marketing accuracy of the merchants. Based on the above access pattern mining technology, this paper analyzes the customer classification support framework of Web mining pattern in E-commerce environment in detail, and discusses the customer management problem in E-commerce environment. In order to correctly formulate enterprise marketing strategy, to help enterprises attract customers, improve profitability.
【學(xué)位授予單位】：天津大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TP393.09;TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前5條

1 郭巖,白碩,楊志峰,張凱;網(wǎng)絡(luò)日志規(guī)模分析和用戶興趣挖掘[J];計算機(jī)學(xué)報;2005年09期

2 張慧穎 ,梁偉;Web使用挖掘中的數(shù)據(jù)預(yù)處理算法研究[J];微型機(jī)與應(yīng)用;2004年08期

3 鮑玉斌,王大玲,于戈;關(guān)聯(lián)規(guī)則和聚類分析在個性化推薦中的應(yīng)用[J];東北大學(xué)學(xué)報;2003年12期

4 趙亮,胡乃靜,張守志;個性化推薦算法設(shè)計[J];計算機(jī)研究與發(fā)展;2002年08期

5 邢東山,沈鈞毅;Web使用挖掘的數(shù)據(jù)采集[J];計算機(jī)工程;2002年01期

，

本文編號：1684972