Microsoft just comes out with a new paper about web ranking at SIGIR (Special Interest Group on Information Retrieval) in Singapore.
If you are curios to read the paper just go here
http://research.microsoft.com/users/tyliu/files/fp032-Liu.pdf and get it !
The idea behind of this paper is simple: Voting and visits. If you get visits and votes then you will get a higher rank. The idea is not so bad, but I don't understand how they will implement that safely. In my opinion they have 3 options:
1. To return to the search page and vote the link what you just visit
2. Integrate in Internet Explorer a voting mechanism (like stumbleupon plugin)
3. All the pages are loading with in a frame and the top frame contains the voting mechanism.
4. Microsoft employ some people to do that.
Let's comment out this implementations:
1. If the user find the information what it looks after, is a very big chance to never return to search engine and vote it.
2. Is a very big chance for lot's of users to never vote this because they don't care about that.
3. This is not a viable solution because then MS can track your every movement.
4. Is a waste of time and resources.
From the paper:
The reasons that BrowseRank outperforms
the other algorithms are as follows:
1) Creating inlinks, which can hurt PageRank, cannot hurt BrowseRank
so much, because the link information is not used in BrowseRank.
2) The performance of TrustRank can be aected by the selection
of the seed set and the determination of the seed distribution in the
link graph. For BrowseRank, seed selection and seed distribution
determination are not necessary.
All the algorithms have the same big problem: If the people know how is acting then can be influenced by people even if the algorithm is a smart one.
Authors of BroseRank paper are Bin Gao, Tie-Yan Liu, and Hang Li from Microsoft Research Asia and Shuyuan He of Peking University, and Ying Zhang of Nankai University, Zhiming Ma of the Chinese Academy of Sciences.