Main research lines
Semiotic Dynamics
A new
paradigm has been quickly gaining ground in information systems
on the World Wide Web: Collaborative Tagging. In web-based
applications like Del.icio.us, Flickr, CiteULike,
BibSonomy, users enrich diverse
resources, ranging from photographs to scientific
references and web pages, with semantically meaningful information in
the form of text labels, or tags. Tags are freely chosen and
users associate resources with them in a totally uncoordinated
fashion, for their own use. Despite its intrinsic anarchist nature,
the dynamics of this terminology system spontaneously leads to
patterns of terminology common to the whole community or to subgroups
of it. Surprisingly, this emergent and evolving semiotic system
provides a very efficient navigation system through a large, complex
and heterogeneous sea of information.
My research is aimed at giving
a scientific foundation to these developments, so contributing to the
growth of the new field of Semiotic Dynamics. Semiotic Dynamics
studies how semiotic relations can originate, spread, and evolve over
time in populations, by combining recent advances in linguistics and
cognitive science with methodological and theoretical tools of complex
systems and computer science.
My research aims at exploiting the unique opportunity offered by the
availability of enormous amount of data. This goal will be achieved
through: (a) a systematic and rigorous gathering of data;
(b) designing and implementing innovative tools and procedures for data
analysis and mining; (c) constructing suitable modeling schemes which
will be implemented
in extensive numerical simulations. My research aims in this way at
providing a virtuous feedback
between data collection, analysis, modeling, simulations and (whenever
possible) theoretical constructions, with the final goal to
understand, predict and control the Semiotic Dynamics of on line
social systems.
Useful links:
- TAGora project: a STREP project funded by the European Commission in the framework of the FET proactive initiative Simulating Emergent Properties in Complex Systems.
Selected articles:
- Vocabulary growth in collaborative tagging systems,
C Cattuto, A Baldassarri, VDP Servedio, V Loreto, arXiv:0704.3316; - A Yule-Simon process with memory,
C Cattuto, V Loreto, VDP Servedio, arXiv:cond-mat/0608672.
Complex Networks
In the last years, much attention has focused on the study of
complex networks.
A network is a mathematical object consisting of a
collection of vertices (nodes) connected by edges
(links).
Networks arise in many areas of science: biology, social
sciences,
Internet, WWW, etc., where vertices and
links can be for example, proteins and their mutual interaction,
individuals and sexual relationship, computers and
cable connections.
Very interestingly the same non trivial statistical properties appear
ubiquitously in all the above situations.
A more traditional view, indeed, is represented by the binomial model
inspired to the random graph model of
Erdös-Rényi, where
each vertex has the same
probability to connect to any other, resulting in a network
with vertex degree, i.e. the number of edges connected to each
vertex, distributed according to a binomial probability distribution.
This is not the case of real data, where instead, the
structure is self similar resulting in a scale-free probability
distribution for the degree.
More specifically, the degree of the vertices is distributed
according to a power law with exponents usually between -3 and -2.
Useful links:
- COSIN project (COevolution and Self-organization In dynamical Networks): a FET OPEN Project, funded by EU commission in the priority area of Information Society Technologies
Selected articles:
- Detecting communities in large networks,
A Capocci, VDP Servedio, G Caldarelli, F Colaiori, arXiv:cond-mat/0402499; - On the Widespread Occurrence of the Inverse Square
Distribution in Social Sciences and Taxonomy,
G Caldarelli, C Caretta Cartozo, P De Los Rios, VDP Servedio arXiv:cond-mat/0311486.
Surface Electronic Structure Calculations
The electronic properties of solids are mainly determined by their
electrons in the bulk. However, in certain circumstances the electrons
close to the surface acquire lot of importance. For example, scanning
tunneling spectroscopy (STS) measurements or photo-emission
(PE)
experiments are able to probe only a narrow region of few layers below
the surface (because of the low penetration depth of electrons).
In order to understand the results of such experiments, we have to
calculate the electronic stucture of solids near their surface.
This task is not simple, since the Bloch theorem is no more valid
along the direction perpendicular to the surface and special tecniques
have to be used to cope with this nasty loss of symmetry.
Useful links:
- IOMP Dresden
- IFW Dresden
- FPLO ( full-potential local-orbital minimum-basis code to solve the Kohn-Sham equations on a regular lattice using the local spin density approximation (LSDA))
Selected articles:
- Photoemission study of the spin-density wave state in thin
films of Cr,
F Schiller, DV Vyalikh, VDP Servedio, SL Molodtsov, arXiv:cond-mat/0307196; - Surface states and their possible role in the
superconductivity of MgB2,
VDP Servedio, S-L Drechsler, T Mishonov, arXiv:cond-mat/0111434.