Web Mining - An Important Tool for Molding Business

By

Usha P.M.
Faculty Member (IT)
ICFAI National College
Kozhikode
 

ABSTRACT

Internet has become the platform for business.  Reasons are varied which includes global market, increased sales and decreased cost. On a global basis, there is a forty percentage increase in the market for online shopping in the last two years (from Nielsen Global Online Survey).  Customer satisfaction is the key secret of success for all industries regardless of whether it is web enabled or not. This article highlights the role of web mining in achieving a viable edge in business.

INTRODUCTION

Web mining is becoming the tool for success for those who adopt electronic means of operation for conducting their business. Web mining is the application of data mining techniques to discover patterns from the Web through content mining, structure mining, and usage mining. Web mining can contribute to a large extent in gaining a competitive advantage in your business. Your business goals should be well understood.  We can explore areas where web mining can help you achieve these goals.

CUSTOMER BEHAVIOR ANALYSIS

Customer relationship management is one of the major applications of Web mining. A website should be designed to entice the customers. Web Mining analyses visitor's behavior and makes predictions on their future interaction. This can be exploited to improve website performance and to recommend products or links based on user's behavior. Visitors entering the site exhibits different behavior. They might just surf through or the process might end up in a purchase. For understanding customer behavior and thus improve the performance of your web site, certain standards should be used. Web metrics provide a method to evaluate the performance. There are certain standard notations used which is provided in the table below.

Information about the visitor's behavior can be collected from Web logs which includes various log files like

We can also analyze the pattern in which the user travelled from one web page to another of your website before making a purchase. Behavior can be analyzed to form user clusters who require same class of details. Information mined can be used to opt for the finest manner of interaction with customers. The decision made by a customer depends on his experience on the site. Provision can be provided to enter feedback. Data can be gathered and mined from web to achieve the final goal of turning visitors to buyers.

WEB MINING IN E-LEARNING

E-learning is nothing but where learning happens with the aid of digital tools and content, with necessary interactions possible. E-learning is independent of location, time or space. A user who is exploring the web for self learning may not be a domain expert. Web sites and topics popular among other users exploring the same area can be found using web mining and recommended to the non-experts. Students or learners using E-learning sites will be differing from each other on various aspects like the rate at which they capture the lessons, domain expertise, personal interests etc. On the other hand, there will be many learners who show the same characteristics also. Interest of learners and relevance of topics to them can be mined from the user's profile created and the same can be made in designing and linking sites.

Giving the right combination of keywords is imperative for searching and finding what we require exactly. Query keywords can be grouped into clusters and information from the query result can be recommended to users. E-learning experience can be enhanced by providing the relevant information to the relevant learner with the assistance of Web mining.

WEB MINING IN BANKING INDUSTRY

 Internet banking was offered as a "value addition" for most customers. But now almost all the banks have ventured into this area. Enormous amount of data gets stored through banking transactions. Success factor is the amount of valuable knowledge that is extracted from this data store. Customer profile can be generated. This helps the bank executives in identifying the appropriate customer for certain category of products and the risk in allotting loan facilities. Credit card usage patterns can be identified and special offers can be provided. Defaulters of payment can be identified easily. Banks like ICICI bank and HSBC bank identify the customers for certain offers like home equity loan using web mining. Bank can target at the customers who are likely to invest in mutual funds. Thus targeting the right customer and ensuring excellent service to these customers can be made a reality with the aid of web mining.

LIMITATIONS

Raw Web data will not be in a presentable form for analysis. Web data will have to be prepared for analysis. The software vendors providing Web Analysis claim of having sophisticated methods for data preparation. When you consider the data available for mining in log files on server side and client side, server records the date/time, bytes transferred, time taken for transaction etc. Client side counts every single visit and page view to a website.  The server log may not contain correct usage pattern because of caching. Request for details from web sites which are already traversed may be served by web cache. This can be avoided by cache busting where browsers should be requesting for fresh copies at every user request.

Another problem is where responders are not genuine. For application like credit offers, frequent responses are not a positive sign. Introduction of incentives and other offers for attracting a customer will result in increase of cost of acquiring a new customer.

CONCLUSION

Web mining is an excellent tool for acquiring an enviable position in business and for sustaining and improving it. But the rate of success depends on the alignment of the outcome of web mining with the strategic goals of your business. Web mining has made its presence felt in the area of e-tailing, e-learning, e-hr and e-finance to list a few. Organizations moving along with these advantages provided by technology had always sailed to success.

References

1. http://www.ebusinessnews.info/?action=read&article=677

2. http://www.webanalyticsassociation.org/en/articles/printview.asp?467

3. http://www.tdan.com/view-articles/4972

4. http://ceodaily.wordpress.com/2008/02/20/can-you-juggle-with-web-metrics/
 

Usha P.M.
Faculty Member (IT)
ICFAI National College
Kozhikode
 

Source: E-mail March 18, 2008

Articles No. 1-99 / Articles No. 100-199 / Articles No. 200-299 / Articles No. 300-399 /
Articles No. 400-499 / Articles No. 500-599 / Articles No. 600-699 / Back to Articles 700 Onward
Faculty Column Main Page