Associate Professor in Complex Systems

Department of Biomedical Engineering and Computational Science (BECS)

Aalto University, Espoo, Finland

&

Research Leader

Complex Networks and Systems Lagrange Lab

Institute for Scientific Interchange (ISI) 

Torino (Italy)

News

• On November the 16th, 2011, I have given my Installation Lecture at Aalto University, a short pedagogical presentation of the history and principles of sociophysics. The video of the lecture can be seen here. 
• I am the winner of the Young Scientist Award for Socio- and Econophysics 2011. The award ceremony was held at the annual meeting of the German Physical Society in Dresden, on March 14th, 2011 (photos: 1, 2, 3, 4). Following the reception of the award, two articles on myself and my work appeared in the newspaper Italia Oggi and in the popular magazine L'Espresso.
  

Recent papers
                                

• In the paper Characterizing and modeling citation dynamics (PLoS One 6, e24926, 2011) we find that citation distributions of networks of papers of the Americal Physical Society are described by shifted power laws and that citation dynamics is characterized by bursts in the early life of papers. Both features can be accounted for by a modified version of linear preferential attachment, where the paper attractiveness is heterogeneously distributed and decays with time.
  
• How do Nobel Prize Laureates accrue their scientific reputation? In the paper How citation boosts promote scientific paradigm shifts and Nobel Prizes (PLoS One 6, e18975, 2011) we show that groundbreaking discoveries make previous publications of the author visible and cited, even in different topics. This "authority effect" is measurable and can be used to distinguish outstanding scholars from normal ones. See the feature by Philip Ball on Nature News!
  
• In the paper Finding statistically significant communities in networks (PLoS One 6, e18961, 2011) we present OSLOM, the first multi-purpose method to find communities in graphs, accounting for edge directions, edge weights, overlapping communities, hierarchies and community dynamics.
  
• What are the principles behind the dynamics of online popularity? In the paper Characterizing and modeling the dynamics of online popularity (Phys. Rev. Lett. 105, 158701, 2010) we show that popularity in online media does not change smoothly, but it experiences wild fluctuations, following power law distributions, due to exogenous factors.
  
• What do communities in real networks look like? In the paper Characterizing the community structure of complex networks (PLoS One 5, e11976, 2010) we have made a systematic analysis of various networked datasets, finding that communities can be categorized in classes according to their statistical properties and that each class comprises networks of the same (or similar) origin (communication, information, biological, social and technological networks).
  
• What is the best algorithm to find communities in networks? In the paper Community detection algorithms: a comparative analysis (Phys. Rev. E 80, 056117, 2009) we have tested the performances of several graph clustering methods on the benchmark graphs we have recently introduced. As a result, the Infomap method by Rosvall and Bergstrom appears to be the most reliable and should be adopted as a first approach, especially when one has no specific information on the network at study.
  
• The Website http://www.physauthorsrank.org/ is a portal that computes rankings between physicists, based on the SARA score, a measure of credit defined through the network of citations between scientific authors. The SARA score is described in detail in the paper Diffusion of scientific credits and the ranking of scientists, Physical Review E 80, 056103 (2009). Featured by Physics!
• In the paper Explosive percolation in scale-free networks (Phys. Rev. Lett. 103, 168701, 2009) we study scale-free networks built with a special process introduced by Achlioptas et al., in which links are placed such to slow down the formation of large clusters. We find that the percolation transition leading to the formation of the giant component displays analytical features at the threshold for any value of the degree exponent λ. For λ>3 the order parameter displays the trivial scaling expected for discontinuous transitions. The percolation threshold is finite for λ>~2.2, in contrast to standard random percolation. In the paper Explosive percolation: A numerical analysis (Phys. Rev. E 81, 036110, 2010) we present a detailed numerical analysis of the process including lattices, random graphs and scale-free networks.
  
• The paper Community detection in graphs (Phys. Rep. 486, 75-174, 2010) is the first comprehensive review article on the problem of graph clustering, which consists in identifying clusters of vertices with a high internal density of edges, whereas the density of edges between clusters is comparatively low.
• The paper Statistical physics of social dynamics (Rev. Mod. Phys. 81, 591-646, 2009) is the first exhaustive review article of the growing field of sociophysics, where complex large-scale social phenomena are described by means of tools and techniques from statistical physics.
  
• In the paper Detecting the overlapping and hierarchical community structure in complex networks (New J. Phys. 11, 033015, 2009) we propose the first method to detect both the hierarchy among the communities and possible overlaps between communities. The method relies on the local optimization of a fitness function.
  
• Is physics more important than biology? In the paper Universality of citation distributions: toward an objective measure of scientific impact (Proc. Natl. Acad. Sci. USA 105, 17268-17272, 2008) we show that the distribution of the number of citations received by a paper, once suitably normalized, is the same in all scientific disciplines, opening the way to a possible objective evaluation of the impact of papers and authors. Read the feature of Philip Ball on Nature News!
  
• In the paper Benchmark graphs for testing community detection algorithms (Phys. Rev. E 78, 046110, 2008) we present a new class of graphs for testing algorithms to find communities in networks. The new graphs are characterized by skewed distributions for the node degrees and the community sizes, which are important features of real graphs, neglected by current benchmarks. The code to build the graphs can be downloaded in the Software section.