In response to a user's query, a search engine generally returns a very large and ranked list of Web pages. Weighted Inter-Cluster Edge Ranking (WICER) is one of the most popular algorithms used for ranking of such lists. It consists of two parameters,  and  , representing inter-cluster and intra-cluster edge weights, respectively. According to WICER, a should be greater than . As these two parameters depend upon the knowledge of users about the topic they are searching for, this paper focuses on finding the percentage of cases in which should be greater than , and vice versa. Research shows that in 21% of the cases, the value of should be greater than , and for the remaining 79%, vice versa.
With the tremendous growth of information, Internet has become an extremely rich and large information warehouse. Because of the size, structured ness and the arbitrariness of releasing of information on the Internet, searching the desired information on the Internet is a crucial task. Search engines are playing an increasing important role in information retrieval by searching query-specific information from the massive pool of WWW.
The list of Web pages returned by search engines is generally very large. As argued by Marchionini (1992), "end users want to achieve their goals with a minimum of cognitive load and a maximum of enjoyment". Correspondingly, a survey done by IProspect and Jupiter Research in the context of studying the behavior of users towards Web searches reveals that 62% of the search engine users click on a search result within the first page of results, and a full 90% of search engine users click on a result within the first three pages of search results. It is also observed that end users tend to type short queries (Anick, 1994; and Croft et al., 1995). Because of this, it is one of the major functions of search engines to make sure that the most important and relevant Web pages related to users query appear top in the list. Thus, the list of results is ranked.
The hyperlink analysis of Web pages has been well adopted in Web community (Botafogo et al., 1992; and Kleinberg, 1999) and various link-based strategies for ranking have been developed in the past for improving Web search query results. The HITS algorithm proposed by Kleinberg (1999) relies on query time processing to deduce the hubs and authorities that exist in a sub graph of the Web consisting of both the results of a query and the local neighborhood of these results. Bharat and Henzinger (1998) augmented the HITS algorithm with content analysis to improve precision for the task of retrieving documents related to users' query. Chakrabarti et al. (1998) made use of HITS algorithm for automatically compiling resource lists for general topics. Sergey Brin and Lawrence Page introduced the Page Rank (Brin and Page, 1998; and Page et al., 1998) algorithm that precomputes a rank vector that provides a priori `importance' estimates for all of the pages in the Web graph. This vector is computed once, is offline, is independent of the users' search query, and has made Google so successful (Page et al., 1998; and Winder, 2005). |
