4 ms·
Toward the Discovery of Citation Cartels in Citation Networks
- nl 10y agoAs someone with a hardly any interest in this problem it seems to me that this isn't a great paper. To discuss this without mentioning the multiple downloadable citation networks or any mention of basic graph analysis techniques (centrality measures or graph cut counts seem relevant here) seems a pretty big oversight. The did mention doping in cycling through.
- SmokyBorbon 10y agoI can tell you neither have any interest nor any knowledge of this field. Someone with experience in bibliometrics, like me, would not embarrass themselves by suggesting a "throw some graph theory at it" approach. https://scholar.google.com/scholar?q=bibliometrics https://scholar.google.com/scholar?q=bibliometrics
- nl 10y agoTo make it clear to the person who didn't understand this already: bibliometrics is a special case of graph theory.
- huac 10y ago> Additionally, every year, many more researchers join the scientific community (e.g., scientists from China and India). Thanks, guys from Slovenia
- parennoob 10y agoYeah, that phrase serves zero actual purpose. That, combined with the sloppy language in the rest of the paper makes me doubt the scientific detachment of these people.
- Uptrenda 10y agoYou also find the same problem on link sharing services (like this one) in that they're extremely easy to game. I have yet to see a paper that proposes a good solution for that kind of thing.
- teyc 10y agoThis is a problem that Google has largely solved. Its original PageRank algorithm was prone to citation networks inflating the value of a page. Nowadays, link farms have lost much of their power.