Our work starts from the definition of an intuitive formula that can be used to order the Web pages according to their importance, showing the need of a modification of this formula on a mathematical basis. Following the thread of this argument we get to a well-founded general formula, that covers many interesting different cases, and among them that of PageRank, the algorithm used by the Google search engine, as it is currently proposed in recent works [4, 7]. Then we prove the substantial equivalence between this PageRank formula and the classic formula proposed in [3]. As an example of the versatility of our general formula we derive from it a version of PageRank based on a user personalization. Finally, we discuss the problem of the "objectivity" of classic PageRank, demonstrating that a certain degree of subjectivity persists, since the order of Web pages given by this algorithm depends on the value of a parameter.
[1]
Sang Ho Lee,et al.
An Improved Computation of the PageRank Algorithm
,
2002,
ECIR.
[2]
Shlomo Moran,et al.
SALSA: the stochastic approach for link-structure analysis
,
2001,
TOIS.
[3]
Rajeev Motwani,et al.
The PageRank Citation Ranking : Bringing Order to the Web
,
1999,
WWW 1999.
[4]
Monika Henzinger,et al.
Hyperlink Analysis for the Web
,
2001,
IEEE Internet Comput..
[5]
Allan Borodin,et al.
Finding authorities and hubs from link structures on the World Wide Web
,
2001,
WWW '01.
[6]
Samuel Karlin,et al.
A First Course on Stochastic Processes
,
1968
.
[7]
Sergey Brin,et al.
The Anatomy of a Large-Scale Hypertextual Web Search Engine
,
1998,
Comput. Networks.