pagerank

18
Motivation When searching for information on the WWW, user perform a query to a search engine. The engine return, as the query’s result, a list of Web sites which usually is a huge set. So the ranking of these web sites is very important. Because much information is contained in the link-structure of the WWW, information such as which pages are linked to others can be used to

Upload: felix

Post on 13-May-2015

433 views

Category:

Technology


0 download

TRANSCRIPT

Page 1: Pagerank

Motivation

When searching for information on the WWW, user perform a query to a search engine. The engine return, as the query’s result, a list of Web sites which usually is a huge set. So the ranking of these web sites is very important. Because much information is contained in the link-structure of the WWW, information such as which pages are linked to others can be used to augment search algorithms.

Page 2: Pagerank

1. The Stochastic Approach for Link-Structure Analysis (SALSA) and the TKC Effect

2. The PageRank Citation Ranking: Bringing Order to the Web

Page 3: Pagerank

Paper 1----SALSA

• authorities: web pages that have many outlinks

• hubs: web pages that point to many authoritative sites

• Hubs and authorities form communities, the most prominent community is called the principal community.

Page 4: Pagerank

SALSA----IdeaSALSA is based upon the theory of Markov chains, and relies on the stochastic properties of random walks performed on our collection of sites.

The input to our scheme consists of a collection ofsites C which is built around a topic t. Intuition suggests that authoritative sites on topic t should bevisible from many sites in the subgraph induced by C. Thus, a random walk on this subgraph will visit t-authorities with high probability.

Page 5: Pagerank

SALSA----Idea

Combine the theory of random walks with the notion of the two distinct types of Web sites, hubs and authorities, and actually analyze two different Markov chains: A chain of hubs and a chain of authorities. Analyzing both chains allows our approach to giveeach Web site two distinct scores, a hub score and an authority score.

Page 6: Pagerank

SALSA----Computing

Now define two stochastic matrices, which are the

transition matrices of the two Markov chains at interest:

1. The hub-matrix H:

2. The authority-matrix à:

Page 7: Pagerank

SALSA

the principal community of authorities(hubs) found by the SALSA will be composed of the sites whose entries in the principal eigenvector of A (H) are the highest.

Page 8: Pagerank

SALSA----Conclusion

SALSA is a new stochastic approach for link structure analysis, which examines random walks on graphs derived from the link structure. The principal community of authorities(hubs) corresponds to the sites that are most frequently visited by the random walk defined by the authority(hub) Markov chain.

Page 9: Pagerank

The PageRank Citation Ranking:The PageRank Citation Ranking:Bringing Order to the WebBringing Order to the Web

Larry Page etc.

Stanford University

Page 10: Pagerank

PageRank----Idea

Every page has some number of forward links(outedges) and backlinks(inedges)

Page 11: Pagerank

PageRank----Idea

Two cases PageRank is interesting:

1. Web pages vary greatly in terms of the number of backlinks they have. For example, the Netscape home page has 62,804 backlinks compared to most pages which have just a few backlinks. Generally, highly linked pages are more “important” than pages with few links.

Page 12: Pagerank

PageRank----Idea

2. Backlinks coming from important pages convey more importance to a page. For example, if a web page has a link off the yahoo home page, it may be just one link but it is a very important one.

A page has high rank if the sum of the ranks of its backlinks is high. This covers both the case when a page has many backlinks and when a page has a few highly ranked backlinks.

Page 13: Pagerank

PageRank----Definition

uBv vN

vRcuR

)()(

u: a web page

Fu: set of pages u points to

Bu: set of pages that point to u

Nu=|Fu|: the number of links from u c: a factor used for normalization

The equation is recursive, but it may be computed by starting with any set of ranks and iterating the computation until it converges.

Page 14: Pagerank

PageRank----definition

A problem with above definition: rank sink

If two web pages point to each other but to no other page, during the iteration, this loop will accumulate rank but never distribute any rank.

Page 15: Pagerank

PageRank----definition

Definition modified:

)()(

)( ucEN

vRcuR

uBv v

E(u) is some vector over the web pages(for example uniform, favorite page etc.) that corresponds to a source of rank.

E(u) is a user designed parameter.

Page 16: Pagerank

PageRank----Random Surfer Model The definition corresponds to the probability

distribution of a random walk on the web graphs.

E(u) can be thought as the random surfer gets bored periodically and jumps to a different page and not kept in a loop forever.

Page 17: Pagerank

PageRank----Conclution

PageRank is a global ranking based on the web's graph structure

PageRank use backlinks information to bring order to the web

PageRank can be thought of as random surfer model.

Page 18: Pagerank

Compare----SALSA and PageRank

Both ranking web page by link structure information.

Both are based on the graph of the web.

Both use random walk idea.