05-03-2011, 04:08 PM
[attachment=9626]
1. INTRODUCTION
Web:
Now a day the internet, otherwise called the web, is widely used in several fields. The advent of e-Commerce has revolutionized several business industries such as banking and insurance.
You can get information studies and communicate with educational institutions. Because of the dynamic nature of the web, you may face challenges to locate any specific domain of information.
Mining:
Mining is a technique of information or knowledge discovery from millions of sources across the Web.
Web Mining:
Web mining is a specialized application of data mining. The web mining is a technique to process information available on web and search for useful data. Web mining enables you to discover web pages, text documents, multimedia files, images and other types of resources from web. Web mining is widely used in several fields.
2. WEB MINING APPLICATIONS
The various fields where web mining is applied are:
E-Commerce
Information filtering
Fraud detection
Education and research
E-Commerce:
In e-commerce, web mining helps in generating user profiles by customizing the choice of users. For example, web mining enables a user to search for an advertisement and information regarding a product of his interest. Internet advertising is one of the major fields in e-commerce, where web mining is widely used. Advertising in a specific domain of an e-commerce web site or a general web site and is considered as one of the major application area of web mining.
Information filtering:
Information filtering is the method to identify the most important results from a list of discovered frequent set of data items for which you can make use of web mining.
Fraud detection:
Fraud detection can be performed using web mining by maintaining a list of signatures of all the users. Web mining is also applied for plagiarism detection and research works.
3. WEB MINING TASKS
The general techniques and algorithms of data mining are also applicable in web mining. Web mining tasks can be decomposed into four subtasks:
Resource Searching: Indicates the task of retrieving documents from the web.
Information selection: denotes automatic extraction of information from the web documents. Several web mining tools such as Web Miner are available to perform this task.
Generalization of patterns: Denotes automatic discovery of patterns across multiple web sites.
Analysis of web documents: Denotes validation and analysis of the extracted patterns.
4. WEB MINING ISSUES
Several software’s are available to perform the various tasks of web mining. For example, you can download a complete web site with its structure using software such as Teleport Pro, Back Street Browser and Grab-a-Site. These softwares are called offline browsers. Offline browsers enable you to download a copy of the web pages of a web site and explore the web pages offline. While browsing the downloaded web pages offline, you need not worry about communication page link and cost of using the web.
An offline web browser can save all the files of a web site in a single folder. In this case, you will lose the path information of a file.
If there are two files with the same name, the previous file will be overwritten. Otherwise, you can save all the files with the original directory structure. This is called saving a mirror of a web site. In such a case, the home page of the web site is saved at the topmost level of the hierarchy of the downloaded files and directories. The directory containing the home page of the web site is called virtual root directory.
Various web analyzing softwares are available to analyse the statistics regarding the range of factors of a web site such as average number of web pages viewed by the visitors. You can also analyse the behaviour of the visitors of the web site by using the various web analyzing softwares. The visitor information is usually stored in a log file in the server of a web site. The structure of a log file varies depending upon the type of web server. The structure of log file of an Internet Information Server (IIS) is different from Apache web server. The web analyzing softwares read data regarding a web site from the log file.
The technique to retrieve visitor based information from web servers based log files and apply this information to analyse data in known as web log mining. For example, consider that a web site consists of four web pages. Whenever you visit the web site, the home page is shown displayed by default. You can access the remaining three pages only by clicking the specific page link provided in the home page.
There are two major types of log files used in web log mining.
1. access log.
2. agent log files.
An access log file maintains a list of all the web pages that the visitors have requested. These web pages include the HTML files and their Embedded graphic images and any other associated files such as texts. An agent log file consists of information about the browser that was used to explore the various web pages.
You can use a web analyzing software to analyse, which web page has been accessed the most by the visitors. Thus, you can verify the behaviour of the visitors that whether the visitors are analyzing softwares are OneStat, Web analizer and Click Tracks Analyzer. The Web analizer works on Linux platform while other web analyzing softwares work on Windows platform.
Creating and maintaining a community-specific web site is another important issue in web mining. People with the same interest can form a community for cooperation and sharing of web documents. For example, some universities and research institutes forms a community. In the web site of each of the community, you can find hyperlinks to other members of the community. This helps users to navigate between different community-specific web sites to share information.
In terms of mathematic, a community is a set of web pages, where each page has multiple links to other members in the community, rather than links outside the community. You can visualize a community in terms of a graph. There can be two types of entities in a community: hub and authority. Hubs do not provide any information, rather hubs points to the sources of information. In other words, hubs specify the location where you can search for information. Examples of a hub are directories and online yellow pages. A single person or a group of persons are authorized to provide information about the concerned subject. These authorized persons are also responsible to update the information at different intervals. A good hub points to numerous authorities. On the other hand, a good authority is linked to by several hubs.
1. INTRODUCTION
Web:
Now a day the internet, otherwise called the web, is widely used in several fields. The advent of e-Commerce has revolutionized several business industries such as banking and insurance.
You can get information studies and communicate with educational institutions. Because of the dynamic nature of the web, you may face challenges to locate any specific domain of information.
Mining:
Mining is a technique of information or knowledge discovery from millions of sources across the Web.
Web Mining:
Web mining is a specialized application of data mining. The web mining is a technique to process information available on web and search for useful data. Web mining enables you to discover web pages, text documents, multimedia files, images and other types of resources from web. Web mining is widely used in several fields.
2. WEB MINING APPLICATIONS
The various fields where web mining is applied are:
E-Commerce
Information filtering
Fraud detection
Education and research
E-Commerce:
In e-commerce, web mining helps in generating user profiles by customizing the choice of users. For example, web mining enables a user to search for an advertisement and information regarding a product of his interest. Internet advertising is one of the major fields in e-commerce, where web mining is widely used. Advertising in a specific domain of an e-commerce web site or a general web site and is considered as one of the major application area of web mining.
Information filtering:
Information filtering is the method to identify the most important results from a list of discovered frequent set of data items for which you can make use of web mining.
Fraud detection:
Fraud detection can be performed using web mining by maintaining a list of signatures of all the users. Web mining is also applied for plagiarism detection and research works.
3. WEB MINING TASKS
The general techniques and algorithms of data mining are also applicable in web mining. Web mining tasks can be decomposed into four subtasks:
Resource Searching: Indicates the task of retrieving documents from the web.
Information selection: denotes automatic extraction of information from the web documents. Several web mining tools such as Web Miner are available to perform this task.
Generalization of patterns: Denotes automatic discovery of patterns across multiple web sites.
Analysis of web documents: Denotes validation and analysis of the extracted patterns.
4. WEB MINING ISSUES
Several software’s are available to perform the various tasks of web mining. For example, you can download a complete web site with its structure using software such as Teleport Pro, Back Street Browser and Grab-a-Site. These softwares are called offline browsers. Offline browsers enable you to download a copy of the web pages of a web site and explore the web pages offline. While browsing the downloaded web pages offline, you need not worry about communication page link and cost of using the web.
An offline web browser can save all the files of a web site in a single folder. In this case, you will lose the path information of a file.
If there are two files with the same name, the previous file will be overwritten. Otherwise, you can save all the files with the original directory structure. This is called saving a mirror of a web site. In such a case, the home page of the web site is saved at the topmost level of the hierarchy of the downloaded files and directories. The directory containing the home page of the web site is called virtual root directory.
Various web analyzing softwares are available to analyse the statistics regarding the range of factors of a web site such as average number of web pages viewed by the visitors. You can also analyse the behaviour of the visitors of the web site by using the various web analyzing softwares. The visitor information is usually stored in a log file in the server of a web site. The structure of a log file varies depending upon the type of web server. The structure of log file of an Internet Information Server (IIS) is different from Apache web server. The web analyzing softwares read data regarding a web site from the log file.
The technique to retrieve visitor based information from web servers based log files and apply this information to analyse data in known as web log mining. For example, consider that a web site consists of four web pages. Whenever you visit the web site, the home page is shown displayed by default. You can access the remaining three pages only by clicking the specific page link provided in the home page.
There are two major types of log files used in web log mining.
1. access log.
2. agent log files.
An access log file maintains a list of all the web pages that the visitors have requested. These web pages include the HTML files and their Embedded graphic images and any other associated files such as texts. An agent log file consists of information about the browser that was used to explore the various web pages.
You can use a web analyzing software to analyse, which web page has been accessed the most by the visitors. Thus, you can verify the behaviour of the visitors that whether the visitors are analyzing softwares are OneStat, Web analizer and Click Tracks Analyzer. The Web analizer works on Linux platform while other web analyzing softwares work on Windows platform.
Creating and maintaining a community-specific web site is another important issue in web mining. People with the same interest can form a community for cooperation and sharing of web documents. For example, some universities and research institutes forms a community. In the web site of each of the community, you can find hyperlinks to other members of the community. This helps users to navigate between different community-specific web sites to share information.
In terms of mathematic, a community is a set of web pages, where each page has multiple links to other members in the community, rather than links outside the community. You can visualize a community in terms of a graph. There can be two types of entities in a community: hub and authority. Hubs do not provide any information, rather hubs points to the sources of information. In other words, hubs specify the location where you can search for information. Examples of a hub are directories and online yellow pages. A single person or a group of persons are authorized to provide information about the concerned subject. These authorized persons are also responsible to update the information at different intervals. A good hub points to numerous authorities. On the other hand, a good authority is linked to by several hubs.
