【摘要】：B2C電子商務(wù)網(wǎng)站系統(tǒng)每天都會(huì)生成大量的產(chǎn)品交易數(shù)據(jù)和訪問日志數(shù)據(jù),這些數(shù)據(jù)中蘊(yùn)含著大量有價(jià)值的信息,例如,訂單的來(lái)源、客戶的行為、訪問者的興趣等。分析這些數(shù)據(jù),不僅能幫助企業(yè)決策層指導(dǎo)B2C電子商務(wù)網(wǎng)站的運(yùn)營(yíng),吸引更多用戶,提高銷售利潤(rùn)還能夠反映出企業(yè)在市場(chǎng)營(yíng)銷、銷售推廣、售后服務(wù)和財(cái)務(wù)管理等各個(gè)方面的狀況�？傊�,對(duì)這些數(shù)據(jù)進(jìn)行深入有效的分析能夠幫助管理者改善客戶關(guān)系、提高企業(yè)各個(gè)方面的競(jìng)爭(zhēng)力。本文在描述了數(shù)據(jù)倉(cāng)庫(kù)的相關(guān)理論,包括數(shù)據(jù)倉(cāng)庫(kù)的概念、基本特征、系統(tǒng)結(jié)構(gòu),B2C電子商務(wù)的概念、OLAP多維數(shù)據(jù)分析基礎(chǔ)上,提出了一個(gè)較完善B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)模型,論文主要工作包含以下內(nèi)容： 1.分析B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)的用戶需求,提出B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)多層次概念模型,設(shè)計(jì)了相關(guān)的維度模型和事實(shí)集,基于該模型完成了部分維度表和事實(shí)表的物理設(shè)計(jì)。 2.分析B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)數(shù)據(jù)源,討論半結(jié)構(gòu)化的數(shù)據(jù)源處理,針對(duì)Web訪問日志,結(jié)合半結(jié)構(gòu)化數(shù)據(jù)的預(yù)處理方法,提出了改進(jìn)的頁(yè)面媒體類型時(shí)間閾值會(huì)話識(shí)別算法。通過對(duì)不同的URL頁(yè)面類型,采用不同的頁(yè)面時(shí)間閾值計(jì)算方法。相對(duì)于已有的對(duì)所有用戶訪問頁(yè)面使用單一的先驗(yàn)閾值和現(xiàn)有動(dòng)態(tài)閾值計(jì)算,該方法能夠更真實(shí)的反映用戶會(huì)話的情況,且識(shí)別的準(zhǔn)確率有了較大提高,為后續(xù)的分析提供高效準(zhǔn)確的數(shù)據(jù)。 3.基于本文提出的B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)模型,構(gòu)建一個(gè)實(shí)驗(yàn)性B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)項(xiàng)目。以Zen Cart網(wǎng)站系統(tǒng)為例,確定分析主題并基于多維建模思想建立各種不同粒度的維度、數(shù)據(jù)集市,并設(shè)計(jì)一個(gè)ETL體系結(jié)構(gòu),包括ETL調(diào)度方案,數(shù)據(jù)預(yù)處理方法等。最后對(duì)訂單數(shù)據(jù)進(jìn)行聯(lián)機(jī)分析,展示B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)的價(jià)值。 本文提出的B2C電子商務(wù)數(shù)據(jù)倉(cāng)庫(kù)模型有以下特點(diǎn)： 1.該模型有針對(duì)性、實(shí)用性較強(qiáng)的特色,它涉及對(duì)內(nèi)、外電子商務(wù)貿(mào)易活動(dòng)中企業(yè)的各個(gè)主要方面,包括頁(yè)面點(diǎn)擊、產(chǎn)品銷售、訂單、用戶對(duì)產(chǎn)品的評(píng)論、銷售利潤(rùn)、倉(cāng)庫(kù)、訂單產(chǎn)品、物流配送等。 2.該模型采用多層次維度設(shè)計(jì),通過合理有效的概念分層,為企業(yè)決策層提供更好的分析視角。最后,本文通過實(shí)驗(yàn)驗(yàn)證了模型的有效性。
[Abstract]:B2C e-commerce website system generates a large amount of product transaction data and access log data every day, which contains a lot of valuable information, such as the source of orders, the behavior of customers, the interest of visitors and so on. The analysis of these data can not only help the decision makers to guide the operation of B2C e-commerce website, attract more users, but also can reflect the marketing and sales promotion of enterprises. After-sales service and financial management and other aspects of the situation. In a word, in-depth and effective analysis of these data can help managers to improve customer relations and enhance the competitiveness of all aspects of the enterprise. On the basis of describing the relevant theories of data warehouse, including the concept, basic characteristics, system structure, concept of B2C electronic commerce and OLAP multidimensional data analysis, this paper puts forward a perfect data warehouse model of B2C electronic commerce. The main work of this paper is as follows: 1. Based on the analysis of user requirements of B2C e-commerce data warehouse, a multi-level conceptual model of B2C e-commerce data warehouse is proposed, and the related dimension model and fact set are designed. Based on the model, the physical design of some dimension tables and fact tables is completed. 2. The data source of B2C e-commerce data warehouse is analyzed and the semi-structured data source processing is discussed. An improved session recognition algorithm of page media type time threshold is proposed for Web access log combined with the pre-processing method of semi-structured data. Through different URL page types, different page time threshold calculation method is adopted. Compared with the existing user access pages using a single prior threshold and the existing dynamic threshold calculation, this method can more truly reflect the user session, and the recognition accuracy has been greatly improved. Provide efficient and accurate data for subsequent analysis. 3. Based on the B2C e-commerce data warehouse model proposed in this paper, an experimental B2C e-commerce data warehouse project is constructed. Taking the Zen Cart website system as an example, the analysis topic is determined and based on the idea of multidimensional modeling, different grained dimensions, data marts are established, and a ETL architecture is designed, including ETL scheduling scheme, data preprocessing method and so on. Finally, the online analysis of order data is carried out to show the value of B 2 C e-commerce data warehouse. The B2C electronic commerce data warehouse model proposed in this paper has the following characteristics: 1. The model has the characteristics of pertinence and practicability. It involves all the main aspects of the enterprise in both internal and external e-commerce trade activities, including page clicks, product sales, orders, users' comments on products, sales profits, warehouses, etc. Order products, logistics distribution, etc. 2. The model adopts multi-level dimension design and provides a better perspective for enterprise decision making through rational and effective conceptual stratification. Finally, the validity of the model is verified by experiments.
【學(xué)位授予單位】：廣東工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2012
【分類號(hào)】：TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前7條

1 殷賢亮;張為;;Web使用挖掘中的一種改進(jìn)的會(huì)話識(shí)別方法[J];華中科技大學(xué)學(xué)報(bào)(自然科學(xué)版);2006年07期

2 方元康;胡學(xué)鋼;夏啟壽;;Web日志預(yù)處理中優(yōu)化的會(huì)話識(shí)別方法[J];計(jì)算機(jī)工程;2009年07期

3 李燕;馮博琴;魯曉鋒;;Web日志挖掘中的數(shù)據(jù)預(yù)處理技術(shù)[J];計(jì)算機(jī)工程;2009年22期

4 范純龍;姜宏飛;李華;;利用圖片類日志信息改進(jìn)會(huì)話識(shí)別質(zhì)量[J];計(jì)算機(jī)應(yīng)用;2010年04期

5 楊富華;;網(wǎng)絡(luò)日志預(yù)處理中優(yōu)化的會(huì)話識(shí)別算法[J];計(jì)算機(jī)仿真;2011年04期

6 蔡浩;賈宇波;黃成偉;黃志強(qiáng);;Web日志挖掘中的會(huì)話識(shí)別算法[J];計(jì)算機(jī)工程與設(shè)計(jì);2009年06期

7 周愛武;程博;李孫長(zhǎng);夏松;;Web日志挖掘中的會(huì)話識(shí)別方法[J];計(jì)算機(jī)工程與設(shè)計(jì);2010年05期

相關(guān)碩士學(xué)位論文 前10條

1 皮濤;基于Lucene的面向主題信息搜索系統(tǒng)的關(guān)鍵技術(shù)分析及應(yīng)用[D];武漢理工大學(xué);2011年

2 周慶華;面向電子商務(wù)的數(shù)據(jù)挖掘研究與實(shí)現(xiàn)[D];中國(guó)人民解放軍國(guó)防科學(xué)技術(shù)大學(xué);2002年

3 張開松;基于Web技術(shù)的數(shù)據(jù)倉(cāng)庫(kù)研究與設(shè)計(jì)[D];武漢理工大學(xué);2005年

4 岳志強(qiáng);制造企業(yè)銷售信息處理及分析系統(tǒng)研究[D];大連交通大學(xué);2005年

5 徐益軍;電子商務(wù)公共服務(wù)平臺(tái)下的Web挖掘系統(tǒng)研究[D];天津工業(yè)大學(xué);2006年

6 卜建峰;電子商務(wù)系統(tǒng)中數(shù)據(jù)融合與OLAP的研究與設(shè)計(jì)[D];西北工業(yè)大學(xué);2007年

7 李勇;數(shù)據(jù)挖掘的算法研究及其在Web日志分析中的應(yīng)用[D];長(zhǎng)春理工大學(xué);2008年

8 張弛;開源OLAP技術(shù)在多媒體教學(xué)系統(tǒng)中的應(yīng)用研究[D];北京郵電大學(xué);2009年

9 蔡俊;基于數(shù)據(jù)倉(cāng)庫(kù)的點(diǎn)擊流技術(shù)的研究[D];江蘇大學(xué);2009年

10 王立;OLAP在視頻網(wǎng)站日志分析中的應(yīng)用[D];東華大學(xué);2010年

，

本文編號(hào)：2307936