Web Data Mining Technology Exploration Essay

In the day-to-day learning, working life, we will more or less come into contact with the dissertation, right, the dissertation for all educators, for the improvement of the overall understanding of mankind has an important significance. So do you know how to write a good thesis? The following is my collection of Web data mining technology to explore the paper, for your reference and reference, I hope it can help the friends in need.

Web Data Mining Technology Exploration Paper Part 1

Introduction

Currently, with the development of network technology and the rapid development of database technology, it effectively promotes the business activities from traditional activities to e-commerce change. E-commerce is the use of computer and network technology and long-distance communication technology, to achieve the entire business activities of electronic, digital and networked. The rapid development of Internet-based e-commerce, so that modern enterprises have accumulated a large amount of data, which not only bring more useful information to the enterprise, but also enable other modern business managers to collect a large amount of data in a timely and accurate manner. Access to customers to provide more and better quality of service, become a key factor in the success or failure of e-commerce, and thus by the modern e-commerce operators a high degree of attention, which also puts forward new requirements for computer web data technology, Web data mining technology came into being. It is a kind of new technology that can obtain a large amount of data from the Internet and effectively extract useful information for enterprise decision-makers to analyze and refer to, so as to scientifically and reasonably formulate and adjust the marketing strategy, and provide customers with dynamic, personalized and high-efficiency services. At present, it has become an indispensable and important carrier of e-commerce activities.

Overview of Computer Web Data Mining

1.Origin of Computer Web Data Mining

Computer Web Data Mining is a process of filtering data and information that is useful to you on Web resources. Web data mining is to transfer traditional data mining ideas and methods to Web applications, i.e., to select patterns or hidden data information from existing Web documents and activities that are of interest and useful to you. Computer Web data mining can show its role in multiple fields, has been widely used in database technology, information acquisition technology, statistics, artificial intelligence in machine learning and neural networks and many other aspects, which play a significant role in promoting the change of business activities is the most obvious.

2. Meaning and characteristics of computer Web data mining

(1) the meaning of Web data mining

Web data mining refers to the application of data mining technology in the Web environment, is a data mining technology and WWW technology combined with the generation of a new technology, the comprehensive use of computer language, Internet, artificial intelligence, Statistics, informatics and other areas of technology. Specifically, it is through the full use of the network (Internet), mining user access log files, commodity information, search information, purchase and sale of information, as well as network user registration information and other content, from which to find out the hidden, potentially useful and valuable information, and then finally used for enterprise management and business decision-making.

(2) the characteristics of Web data mining

Computer Web data mining technology has the following characteristics: First, the user does not have to provide subjective evaluation of information; Second, the user "access mode dynamic acquisition" will not be obsolete; Third, can handle large-scale data volume, and easy to use; Fourth, compared with traditional databases and data warehouses. Compared with traditional databases and data warehouses, the Web is a huge, widely distributed, global information service center.

(3) Categories of computerized web data mining techniques

web data mining techniques **** there are three categories: the first category is Web usage record mining. It is to mine Web log records through the network to find the pattern of user access to Web pages and potential customers and other information, in order to improve the competitiveness of all services of its site. The second category is Web content mining. Both refers to the process of extracting knowledge from Web documents. The third category is Web structure mining. That is, through the content of a large number of documents on the Web collection of small summary, clustering, correlation analysis of the way, from the organizational structure of Web documents and linking relationships in the prediction of relevant information and knowledge.

The relationship between computer web data mining technology and e-commerce

With the maturity of computer technology and network technology, e-commerce is fast and convenient by more and more enterprises and individuals. With the continuous expansion of e-commerce business scale, e-commerce business goods and the number of customers also increased rapidly, e-commerce enterprises to obtain a large amount of data, these data are becoming an e-commerce business customer management and sales management of important information. In order to better develop and utilize these data resources so as to bring more convenience and benefits to enterprises and customers, various data mining techniques have been gradually applied to e-commerce websites. At present, the e-commerce recommender system built on data mining (especially web data mining) technology is becoming a trend in the development of e-commerce recommender system.

Specific application of computer web data mining in e-commerce

(1) The process of web data mining in e-commerce

In e-commerce, the process of web data mining mainly has the following three phases: both the data preparation phase, the data mining operation phase, and the results of expression and interpretation phase. If in the result expression stage, the analysis results can not make the decision makers of e-commerce enterprises satisfied, it is necessary to repeat the above process until satisfied.

(2) Web data mining technology in e-commerce

At present, e-commerce is widely used in enterprises, greatly promoting the rise of e-commerce sites, after analyzing the access information of users on the site for a certain period of time, it can be found that the business site on the potential customer groups, related pages, clustering of customers, and other data and information, enterprise information systems. Therefore, a large amount of data will be obtained, so much data to make Web data mining has a rich data base, so that it has a more important in a variety of business areas. Practical value. As a result, e-commerce will be the main direction of Web data mining in the future.The application of Web data mining technology in e-commerce mainly contains the following aspects:

First, find potential customers. E-commerce activities, enterprise sellers can use classification technology to find potential customers on the Internet, by mining Web log records and other information resources, visitors to the classification, looking for access to the customer **** the same characteristics and laws, and then from the classification already exist to find potential customers.

The second is to retain visiting customers. E-commerce enterprises through the business website can fully explore the information left behind when customers browse the visit to understand the customer's browsing behavior, and then according to the customer's different preferences and requirements, and then make timely visits to the customer's satisfaction with the page recommendations and exclusive products, in order to continuously improve the satisfaction of the site visit, maximize the extension of the customer's residence time, to achieve the purpose of retaining the old customers to explore new customers.

Third, to provide marketing strategy reference. Through Web data mining, e-commerce business sellers can dig through the commodity access situation and sales, while combining the market changes, through the clustering analysis method, to deduce the customer access to the law, the different consumer demand and consumption of product life cycle and other situations, to provide timely and accurate information for decision-making reference, so that decision makers can make timely adjustments to the commodity sales strategy, optimize the The first step is to make sure that you have the right tools for the right job.

Fourth, improve the business website design. E-commerce site designers can use the rules of association to understand the customer's behavioral records and feedback, and as a basis for improving the site, and constantly optimize the site's organizational structure to facilitate customer access, and constantly improve the site's click-through rate.

Conclusion

This article provides an overview of Web data mining technology, which is widely used in e-commerce. It can be seen that with the rapid development of computer technology and database technology, the application of computer Web data technology will be more extensive, Web data mining will also become a very important research field, the research prospect is huge and far-reaching. At present, China's Web data applications are still in the exploration and start-up phase, there are many issues worthy of in-depth study.

Web Data Mining Technology Exploration Paper Part 2

Abstract: The paper analyzes the application of WEB data mining technology in e-commerce by introducing the basic knowledge of e-commerce and data mining, respectively from several aspects.

Keywords: e-commerce; data mining; application

1 Overview

E-commerce refers to the activities of enterprises or individuals to use the network as a carrier, the application of electronic means, and the use of modern information technology for the exchange of business data and the conduct of business operations. With the rapid development of the Internet, e-commerce has more obvious advantages than traditional business, due to e-commerce has a convenient, flexible, fast characteristics, so that it has gradually become an indispensable activity in people's lives. At present, there are many e-commerce platform websites and strong competition in the industry. In order to obtain more customer resources, e-commerce websites must strengthen customer relationship management, improve the business philosophy and enhance after-sales service. Data mining is the process of identifying implicit, potentially useful, effective, novel, and understandable information and knowledge from a data set. Inductive reasoning from the data set, from which mining and commercial prediction, can help e-commerce business decision makers based on the prediction, the market strategy adjustment, will reduce the risk of the enterprise, so as to make the right decision, the enterprise profit will be maximized. With the increasingly widespread application of e-commerce, e-commerce activities will generate a large amount of useful data, how can data mining data reference value? Study the interests and hobbies of customers, classify customers into different categories, and recommend the goods that customers desire to the relevant customers respectively. Therefore, how to carry out data mining on the e-commerce platform has become a hot issue in research.

2 Data Mining Technology Overview

DataMining, also known as KnowledgeDiscoveryinDatabase (KDD). Data mining generally refers to the process of applying algorithms to find out hidden and unknown information from massive data. Data mining is a process of using analytical tools to discover the relationship between models and data in big data resources.Data mining plays a key role for decision makers to look for some kind of potential correlation between data and discover hidden factors. These patterns are potentially valuable and can be understood. Data mining will be artificial intelligence, machine learning, databases, statistics, visualization, information retrieval, parallel computing and other fields of theory and technology together in a multidisciplinary cross-disciplinary study, these disciplines also provide a great technical support for data mining.

3 Features of Web Data Mining

Web Data Mining is the application of data mining in the Web.The purpose of Web Data Mining is to find valuable data or information from the content of Web pages on the World Wide Web, the structure of the hyperlinks, and the use of log records. Based on the type of data used in the mining process, Web data mining tasks can be divided into: Web content mining, Web structure mining, Web usage log mining.

1) Web content mining refers to the extraction of text, images, or other information that make up the content of the web page, the mining object usually contains text, graphics, audio and video, multimedia, and various other types of data.

2) Web structure mining is mining the structure between Web pages, mining describes how the content is organized, from the Web's hyperlink structure to find the Web structure and the page structure of valuable patterns. For example, from these links, we can find out which are the important web pages, based on the theme of the web page, automatic clustering and classification, for different purposes from the web page according to the pattern to obtain useful information, so as to improve the quality and efficiency of retrieval.

3) Web Usage Log Mining is a method based on mining the access logs of user accesses on the server.Web Usage Mining maps the log data to a relational table and uses the corresponding data mining techniques to access the log data, and collects and analyzes user clicking events to discover the user's navigational behavior. It is used to extract link information about how customers navigate and use visited pages. Such as which pages were visited? Time spent in each page? What was clicked next? What route was taken to exit the browsing? These are all questions that Web usage record mining is interested in addressing.

4 E-commerce Web mining in the application of technology analysis

1) E-commerce in the application of sequential pattern analysis

Sequential pattern data mining is to mine patterns based on time or other sequences. For example, in a set of chronological sessions or transactions an item has presence following another item. By doing this, WEB sellers can predict future access patterns to help target specific groups of users for advertisement emission settings. Sequential patterns are found to make it easy for customer behavior to be predicted by the e-commerce organizer as users browse the site, catering to each user's browsing habits as much as possible and continually adapting the web pages to the user's interests to satisfy each user as much as possible. Mining logs using sequential pattern analysis reveals the sequential patterns of customer visits. In World Wide Web usage log mining applications, sequential pattern mining can be used to capture common navigation paths among user paths. When a user visits an e-commerce site, the webmaster is able to search for sequential patterns of that visitor's visits to the site and recommend pages that the visitor is interested in but has not yet viewed. Sequence pattern analysis can also analyze the order of goods purchased before and after, so as to recommend to the customer. For example, when a search engine is used to send out a query request, browse web information, etc., advertisements related to such information will pop up. For example, a user who has purchased a printer will generally purchase printing supplies such as printer paper and toner cartridges shortly. A good recommender system will create an exclusive store for the customers and adjust the content of the website by the characteristics of each customer. It can also analyze the effectiveness of website and product promotions by mining some sequential patterns.

2) The application of association rules in e-commerce

Association rules are to reveal the implied interrelationships between the data, and the task of association analysis is to discover the association rules or correlation procedures between things. The goal of association rule mining is to find out the intrinsic relationship of each data information in the data items. Association rule mining is all about searching out the connections between the content, pages, and files that users access on the server to improve e-commerce website design. The site can be better organized to reduce the burden on the user to filter the information on the site, which items will the customer be likely to buy at the same time in one shopping trip? Association rule technology can analyze the shopping habits of customers through the connection between different items in the shopping basket. For example, 90% of customers who buy milk will also buy bread at the same time, which is an association rule that will increase the sales of these two items if the store or e-commerce site sells them together. The goal of association rule mining is to use tools to analyze the connections between the items purchased by customers, i.e., typical shopping basket data analytics applications. Association rules are used to discover the correlation of different items in similar events, for example, cell phone plus rechargeable battery, mouse plus mouse pad and other purchasing habits belong to association analysis. Association rules mining technology can use the corresponding algorithms to find out the association rules, for example, in the above example, the merchant can improve the placement of goods based on the association between the goods, if the customer purchased a cell phone, then the charging treasure will be put into the recommended goods, if some of the goods were purchased at the same time the probability of a larger, that is, there is a correlation between these goods, the merchant can be associated with these goods links are put together to recommend to the customer, and the sales of goods will also be facilitated. Favorable to the sale of goods, merchants are also based on the correlation of the effective collocation of goods, improve the level of commodity management. Such as buying lamps and lanterns customers, most will also buy switches and sockets, therefore, generally will be lamps and switches and sockets and other items in an area for customers to buy. Based on the analysis to find out the association rules of the goods that customers need, by mining analysis results to recommend to customers the required goods, that is, to the customer may be interested in recommending the goods, will greatly improve the sales of goods.

3) The application of path analysis technology in e-commerce

Path analysis technology through the Web server log files in the customer access to the site of the number of visits to analyze the Web site used to find the Web site in the most frequently accessed paths to adjust the site structure to help use the user the fastest possible speed to find the products or information they need. For example, when a user visits a website, if there are many pages that the user is not interested in, it will affect the user's web browsing speed, thus reducing the user's interest in browsing, and will also increase the maintenance cost of the entire site. The use of path analysis technology can comprehensively grasp the association between the various pages of the site as well as the link between the hyperlinks, through the analysis of the highest frequency of access to the page, so as to improve the site structure and page design.

4) Application of classification analysis in e-commerce

Classification techniques play an important role in Web analytics applications that model users according to a variety of predefined rules. For example, given a set of user transactions, one can calculate the sum of purchase records for each user in a given period. Based on this data, a classification model can be built to categorize users into those who are inclined to make purchases and those who are not, taking into account features such as the user's statistical attributes as well as their navigational activities. Classification techniques can be used both to predict which buying customers are interested in which type of promotional tactics and to predict and classify customer categories. In e-commerce, through classification analysis, it is possible to know the interests and purchase intentions of various types of customers, and thus to identify potential customers, so as to provide personalized online services and targeted business activities for each type of customer. Through the classification and positioning model to assist decision makers to locate their best customers and potential customers, improve customer satisfaction and loyalty, maximize customer profitability, in order to reduce costs and increase revenue.

5) Application of cluster analysis in e-commerce

Clustering technology can cluster data items with the same characteristics into a class. Cluster analysis is to compare the relevant data in the database and find out the relationship between the data, and categorize the data with different nature characteristics. The goal of cluster analysis is to collect data to categorize on the basis of similarity. According to have the same or similar customer buying behavior and customer characteristics, the use of cluster analysis techniques to effectively segment the market, after segmentation should be able to each type of market to develop targeted marketing strategies. There are two types of clustering: page clustering and user clustering. User clustering is to establish user groups with the same browsing patterns, which can be used in e-commerce for market segmentation or to provide personalized Web content to users with similar interests, and more analysis based on statistical attributes of users (e.g., age, gender, income, etc.) can be used to discover valuable business intelligence on user groups. A fine-grained differentiation of markets in e-commerce is the use of cluster analysis techniques. Cluster analysis can be based on the customer's purchasing behavior to classify different customer characteristics of different customer groups, through clustering of customers with similar browsing behavior, allowing marketers to classify customer segmentation, able to provide customers with more humane and attentive service. For example, through the clustering technology analysis, found that some customers like to visit the web page about auto parts content, you can dynamically change the site content, so that the network automatically send these customers clusters of new product information or mail about auto parts. Classification and clustering often interact with each other. In e-commerce by clustering customers with similar behaviors or habits, to provide customers with more satisfactory service. Technicians use cluster analysis to first cluster and subdivide the data to be analyzed, and then use classification analysis to classify and label the data set, and then reclassify that labeling, and so on and so forth to obtain relatively satisfactory results from the two methods of analysis.

5 Conclusion

With the rapid development of the Internet, the application of big data analysis is becoming more and more widespread. E-commerce in commercial trade accounts for an increasing proportion of the use of web mining technology for commercial massive data mining processing, analyzing customer buying preferences, tracking market changes, adjusting sales strategies, decision makers to make effective decisions and improve the market competitiveness of enterprises is of great significance.

References:

[1]Pang Yingzhi.Web data mining technology in e-commerce[J]. Intelligence Science,2011,29(2):235-240.

[2]Ma Zongya,Zhang Huiyan.Web Data Mining Technology in E-Commerce[J]. Modern Economic Information,2014(6):23-24.

[3]Xu Jianbin.Web data mining technology in e-commerce[J]. Times Finance,2013(4):234-235.208

[4]Zhou S.D. Research on the application of Web data mining in e-commerce[D]. Beijing Jiaotong University,2008.

[5]Duan Hongying.Web data mining technology in e-commerce[J]. Journal of Longdong College,2009(3):32-34.

;