Vendredi 11 mai 2012 à 14h15
Auditoire Stueckelberg, Ecole de Physique
Statistical Physics of Citations
Sidney Redner, Department of Physics, Boston University
What are the statistical properties of citations to scientific articles? This question will be discussed using recently available large-scale citation data. I will first present amusing empirical facts about citations from all Physical Review publications over the past 120 years. Intriguingly, the evolution of citations appears to be described by linear preferential attachment, and I will discuss the implications of this mechanism on the properties of the citation network. Paradoxically, the citation distribution for the Physical Review appears to be log-normal, which does not accord with linear preferential attachment. I will also discuss the application of the Google page-rank algorithm to find scientific "gems" in Physical Review---ground-breaking publications that are not so well cited. Finally, I will show how major scientific fields and their connections can be uncovered by resolving the citation network into its constituent communities.