Hacker News Reader

Comment by sytelus | Hacker News Reader

It strikes me that you can use faculty publications and PageRank algorithm to assign a score to each faculty member. Then you can actually find out which university has the strongest CS department. You can extend this to may be other fields to rank universities for each major and replace things like Patersons guide and other university ranking systems which are rather arbitrary.

View on HN

I think you may be interested in the "h-index" and related metrics.

http://www.cs.ucla.edu/~palsberg/h-number.html

http://en.wikipedia.org/wiki/H-index

http://researchguides.uic.edu/hindex

I wrote a quick script to generate a list similar to (but not as deep or well-annotated as) Prof Palsberg's.

https://gist.github.com/shazeline/d9881d06be31a59a93d3

It's a BFS for Google Scholar seeded with Herbert Simon.

The h-index seems like much weaker metric than PageRank. For instance, consider 10 professors you decides to game this. They will create 100 fake papers and cite each other in all. Now suddenly they would have h-index of 100!

PageRank is recursive metric in graph. In above case, if above 10 professors don't have any other links pointing to them, they don't get ranked higher. PageRank was designed to eliminate scenarios like above.

It's amusing that Google Scholar doesn't use PageRank prominently instead of these weaker metrics.

Aren't there areas of scientific research where people doing genuinely valuable research don't get cited at all except within a closed community? I'd imagine e.g. that a handful of researchers all studying some obscure animal species would have a high h-index but a very low PageRank.

An experiment that did just that a few years ago:

"Manipulating Google Scholar Citations and Google Scholar Metrics: simple, easy and tempting" by Delgado Lopez-Cozar, et al.

http://arxiv.org/abs/1212.0638

[edit: grammar]

You can even do something similar at the city level and see which city is the most influential in Science (or physics in this case): http://www.nature.com/srep/2013/130410/srep01640/full/srep01...

"weaker" metrics can be useful, because you can understand it better. If I tell you a PageRank of an author is 3.452. That means nothing to me. If I tell you somebody's h-index is 18. That means they published 18 papers that got 18 citations each. The h-index is useful for comparing people in the same field. There is also a metric for journals called impact factor.

http://en.wikipedia.org/wiki/Impact_factor

PageRank was originally designed for this purpose.

Care to elaborate?

I think warrenmar is referring to the fact a paper's impact can be measured by counting the number of times other authors cite the original paper.

IIRC, Larry and Sergey used that idea as the conceptual framework behind PageRank. Indeed, they were right: the relevance of a search result can be increased tremendously by indexing as many pages as possible and then calculating how many people are linking to said page.

They don't do it like that anymore but that's the basic idea.

The original funding Larry got was from the Digital Library Initiative. Then the web blew up. Maybe I shouldn't say PageRank was designed for citations, but that is where it draws its influence from. Things don't magically pop out of no where. But if you look at the PageRank algorithm, it looks like it is designed for paper citations. As a graduate student, you start a some place and you end up at a different place from where you intended to end up.

My point is that one shouldn't be surprised to see that PageRank would work well for paper citations.

http://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=100660

https://en.wikipedia.org/wiki/Page_rank

PageRank evolved from Kleinberg's Hubs and Authorities[1], which partly evolved from work in ranking research papers.

[1] http://en.wikipedia.org/wiki/HITS_algorithm

I wonder what the equivalent of link farms and spam in such a system would be ?

The Information Systems literature.

Self citation and publishing very similar articles multiple times. Both are already problems, since they work well to game many currently used citation metrics.

Reply on news.ycombinator.com