Skip to main content



The Modification of PageRank for Scientific Importance

When Google created the PageRank system, it was probably unprecedented how it would spread wildly for so many different purposes.  What started as a system to measure the importance websites had wide implications, one of which was for the categorization and importance of scientific papers.  Scientific research is, as we’ve covered in class, a giant network of collaborators and interconnectedness.  Thus, it makes sense to model collaborations in the same way.  In the article, the author states that citations are often skewed towards people who published in more famous journals.  However, the number of citations a person have means very little towards scientific contribution.  It’s also very difficult to model citations as different fields value citations in different ways.  Researches have thus tried to model importance of scientific papers based on many factors like number of links to a paper, number of links to a citation of that paper and interconnectedness of the papers.  Researchers have realized that PageRank is a good way to model these papers.

Using PageRank on a series of papers in a network, researchers found that the papers that came up most in importance were papers written by people who are highly renowned.  Nobel Prize winning papers and other highly regarded papers came up.  This clearly was a success to the researchers.  With a few modifications, CiteRank (the research equivalent of PageRank) was created and is so far a very good way for the relevance of scientific papers to be shown.  This is a direct application of the method we discussed in class and shows how people have been able to manipulate PageRank for so many other purposes.  For now, it is unknown how accurately this system represents importance of publications.  As you get to lesser and lesser important publications, it may show less accuracy more frequently and only time will tell exactly how accurately model this situation

Source:

http://arstechnica.com/science/2009/01/using-pagerank-to-assess-scientific-importance/

Comments

Leave a Reply

Blogging Calendar

October 2016
M T W T F S S
 12
3456789
10111213141516
17181920212223
24252627282930
31  

Archives