vitaplease - 12:27 pm on Jul 8, 2003 (gmt 0)

Nice work Claus.

as discussed in the thread Mohamed referenced to, it was more or less acknowledged as being a variation on Kleinberg.

The big question was if such an additional reranking could be done on the fly.

(OldScore ~ Page Rank).

I like your presentation but I would try to avoid referencing Oldscore with Pagerank or PR. OldScore(x) refers to the relevance score value for the particular document. I know you mean the Pagerank algo, but it can get confusing.

Let's assume, as a very simple example only, that values 0,1,2 are the only values for both PR and LR: If You get a PR of 2 and a LR of 0, then the NewRank will be 0. If you get a PR of 0 you will not even make it to the top set of 1000 that will ever get a LR calculated. On the other hand, if you get a PR and a LR of 1 then you're better off than the guy having a top PR but no LR.

NewScore(x) = (a+LocalScore(x)/MaxLS)(b+OldScore(x)/MaxOS)

Ok in the formula they state a and b be constants, so NewScore would most likely not be 0 if either LS or OS are 0, but you are making a good point.

Lets say a guestbook spammer gets high ranking on a query (oldscore - not necessarily Pagerank) only from totally unrelated sources (not in the initial set of e.g. 1000 results). NewScore would then let rankings for this page drop dramatically because of the multiplier effect from the non-existant LocalScore.

What this paper basically says is, for queries you want to rank high for,
focus to get motivated links from pages ranking in the top of the search engine results for that search query.