making a better web searchingmechanism. To accomplish that task, ituses several approaches and analyzingmethods to do the identifications onthe results data (web resources).In 1998, Yahoo used the personalizedweb pages in their searchingmechanism. As a leading websearching service, Google launchesseveral services for personalizing theweb search. Google on Wednesdayintroduced a personalized web searchservice which was executing by storing
users’ search histories and suggesting
the relevant web resources for eachindividual. On March 1, 2006 Googleintroduced a Google Labs projectnamed My Search History for
customizing the users’ interests.
Another Google’s produce, iGoogleretrieves users’ information and
interests and produces a customizedGoogle home page. By improving thefunctionality of iGoogle, Googlesearch engine uses this personalinformation to make user profiles forpersonalizing the search results foreach user .In 2009, there was another servicecalled Google Real Time Searchcombines the frequent updates madeby the users of social community siteslike Facebook, Twitter and MySpace
with Google’s general web search
results .Figure 2.1 personalized web Page
3. Major Researches in PersonalizedWeb search
This section is for describing the factswhich are related with the approaches,researches of Personalized Web Searchconcept. There are many analyzingapproaches and page ranking methodswhich are to be discussed in here andstudies throughout this independentstudy.
3.1 User Profiling
In this approach, separate informationset about users should be maintainedand updated according to the latest
changes of the users’ facts.
Previouslyvisited web pages, total visit time , lastvisit, number of visits, expiration time,text content, bookmarks and number of out-
links(user’ search history)
areconsist of the typical user profile. Inaddition to that, there are client datalike IP addresses, access method.Browser/Operating systems andpreviously search queries in theprofiles. According to these userprofiles, web search engine analysesthem by an appropriate algorithm andproduces a list of web pages as anoutcome.Above mentioned user information canbe divided into two categories. Someinformation can be provided by theuser (explicitly collecting) and othersshould be learnt automatically byactivities of the users. Most of thesearch engines are highly concerned
about gathering users’
historical dataand preferences automatically becauseusers dislike and reluctant to providetheir personal interests.Keeping long term interests in the userprofile becomes wastage of resources
and an effective thing, because users’
interests are not static, it differsaccording to time and circumstances.