2012/08/28 by M. Golosovsky, Michael Golosovsky, Sorin Solomon · 60 citations
Computer Science · Economics, Econometrics and Finance · Mathematics · Physics and Astronomy · #Advanced Thermodynamics and Statistical Mechanics #Citation #Complex Network Analysis Techniques #Complex Systems and Time Series Analysis #Complex network #Computer science #Exponent #Machine learning #Markov chain #Mathematics #Physics #Point process #Preferential attachment #Statistical physics #Statistics #cond-mat.stat-mech #cs.DL #cs.SI #physics.soc-ph #stat.OT
paper · pdf · doi:10.1103/physrevlett.109.098701
published in Physical Review Letters 109(9), 098701 (American Physical Society) · 16 pages, 9 figures
openalex publication_date 2012/08/28 · arxiv created 2012/10/02 · arxiv updated 2013/02/15 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
We put under experimental scrutiny the preferential attachment model that is commonly accepted as a generating mechanism of the scale-free complex networks. To this end we chose a citation network of physics papers and traced the citation history of 40,195 papers published in one year. Contrary to common belief, we find that the citation dynamics of the individual papers follows the superlinear preferential attachment, with the exponent α=1.25-1.3. Moreover, we show that the citation process cannot be described as a memoryless Markov chain since there is a substantial correlation between the present and recent citation rates of a paper. Based on our findings we construct a stochastic growth model of the citation network, perform numerical simulations based on this model and achieve an excellent agreement with the measured citation distributions.