A Hybrid GA-PSO based approach for Mining Top-Ranked Web Pages to Reorganize Websites

There are millions of websites in the world, and these are increasing gradually. Locating pertinent information from the Internet is a very challenging job. Most of the websites consist of several web pages. These web pages may reach from hundreds to thousands in a single website. It is a difficult task for website designer as well as owners of the website to organize these large number of web pages so that website becomes interactive, and navigation in the website becomes easier. At the same time, required information could be easily obtained to the users. It is possible only if we could bank bees web pages based on the parameters like access frequency, unique visitors to a web page, active time spent by the user, common keywords given by the user, hubs, and authority count values. In this paper, the problem of ranking these web pages has been addressed by the hybrid GA-PSO algorithm where the web pages are first ranked by GA algorithm, and thereafter, these ranked optimal web pages are further refined by the PSO algorithm. Experimental evidence has shown that the proposed hybrid approach allows to find out the top-ranked ‘K’ documents to reorganize the structure of website.