1,721,060 research outputs found
Modelling Multiple Interactions in Science and Technology Networks
No full textEmpirical evidence shows that knowledge transfer from research institutions and universities contributes to local innovation processes in a variety of ways. Several studies have also shown that the extent of innovation diffusion is greatly affected by the structure of the network in which innovation processes take place. In this contribution, we aim to identify the complex structure of relationships at the basis of knowledge and innovation diffusion among actors from various organizations (firms, academic and research institutions) in a given territory. A multiplex approach is proposed to explain co-authorship and co-inventions interactions among Author-Inventors community in a specific geographic area. To this end, we carry out a case study of the Trieste area (North-East Italy), characterized by a very high concentration of research organizations and by the emergence of a lively sector of small firms working in the field of R&D
Prototyping and comparing networks through Archetypal Analysis
No full textIn recent years, in different fields, it has become possibile to observe large collections of networks
referring to the same phenomenon, e.g. sets of collaboration networks, each describing different
scientific field; sets of ego networks, where egos belong to the same category; sets of governance
networks; sets of brain networks. Given such sets of networks, it could be of interest the comparison of
net- works among each other. At the same time it could be relevant the detection of a small number of
representative networks that can serve as a condensed view of the entire collection of networks. In this
paper we focus on this latter aim which amounts, in a statistical perspective, in finding what would be
called prototypical networks able to typify the network structures starting from the observed ones. To
this aim, we adopt the approach proposed in Ragozini et. al, (2016) in the framework the analysis of N
statistical statistical units, described by p variables, synthesized by a set m prototypes. The procedure we
propose goes through 3 steps: i) describe a network through a mixture of features referring to local,
global, and intermediate-scale (meso-scale) network structure ii) find in the space of descriptors a set of
prototypes by applying the procedure proposed in Ragozini et. al, (2016); iii) find in the original space of
the networks, on the base of results of the previous steps, the prototypical networks and profile them.
This procedure allows us to typify the most characteristic network structures in the observed set of
networks, and to have prototypical networks that are characterized by clear and interpretable profiles in
terms of their most relevant features and their specificity in contrast to the others. We demonstrate via a
simulation study how the proposed procedure is able to discriminate and describe different types of
networks derived from several generative models
A relational distance based approach to network evolution
No full textThe study of networks evolution has rapidly become a fundamental topic in the social network analysis (SNA) framework (see for instance, Doreian and Stokman, 1997). In very general terms, networks time evolution may be analyzed in two ways: by considering the dynamics of the behavior of the actors/nodes involved within the network or by taking into account the evolution of the network itself over time . The first approach is related to the actor-oriented class of models, in which the actors activate underlying theoretical micro-mechanisms that induce the evolution of social network structures on the macro-level (Snijders 1999, Snijders et al. 2009). The second approach considers network as a system that evolves over time following specific attachment rules (Watts and Strogatz, 1998; Albert and Barabasi, 1999).
In this paper we adopt an intermediate approach. From one hand, as in actor-oriented perspective, network observations are viewed as discrete states of the evolution process and the changes in the structure are operated actor-wise. On the other hand we assume that the actor’s choices are guided by a mechanism based on their “global” relational position in the net.
We suppose that actors’ choices are governed by the evaluation of their relational distances from the others. We use as relational distance among actors the so-called Euclidean Commute-Time Distance (ECTD) (Jagers, Gobel, 1976; Fouss et al.., 2007), based on the use of some spectral graph theory quantities, as the laplacian matrix of a graph and its pseudo-inverse (Chung, 1997; Bollobas, 2001). This distance has a nice re-interpretation, in SNA terms, because includes, in only one measure, several actor positional characteristics.
In our approach, ECTD is used to define a baseline mechanism that explain ties formations when actors do not have any information on the alters except global network information.
The proposed approach starts from an observed network on N nodes at time t=0, G(0)(V,E(0)) (with adjacency matrix A(0)) and, at least another state of the same network detected in a different time G(t)(V,E(t)) (with adjacency matrix A(t)). In order to specify the following configurations of the network along the successive discrete time occasions (i.e. time t=1,2,...,K), we focus on actor possible choices in terms of activation of new links but also on the deactivation of old links. In particular, we select two classes of candidate nodes that, at the time t, have to decide their changes: a class of unconnected nodes and a class of connected nodes. By the evaluation of the actors’ ECTD distances in this two classes, we select the candidate actor that will determine the next step in the network evolution process.
This approach may be useful to furnish a sort of baseline model for more complex network evolution specifications. In particular, if this mechanism of attachment works, it means that actors do not have access to exogenous information on the alters in the network (i.e. actors of a web social network, firms in an open market)
Social Networks Comparison by Using Laplacian Matrix and Euclidean Commute Time Distance
No full textThe presented paper deals with the comparison between two social networks with the same set V of actors. The main purpose is to develop an exploratory strategy to investigate the relational differences (or similarities) between networks. Markov Chain on graph is used in order to introduce a special euclidean node-distance function. This distance is computed by means of the laplacian matrix L. The procedure generates for each of the two networks a distance matrix Δk (k = 1,2). differences are detected by projecting these matrices in a common and more parsimonious Euclidean space obtained by a multidimensional scaling and a procrustes analysis
Gli atteggiamenti verso la diversità: un'analisi dei dati statistici su discriminazione e integrazione nel contesto italiano e regionale
No full textIl tema dell’intolleranza e delle difficoltà di integrazione delle comunità straniere nei territori nazionali è di crescente interesse a livello comunitario. In sede europea si è deciso di attuare dei programmi volti a rendere sistematici ed effettivi il principio di parità di trattamento e non discriminazione.
In questo contesto uno dei principi chiave alla base di tale piano comunitario vi è quello di raccogliere e usare i dati sull’uguaglianza e sui crimini motivati dall'odio per garantire una politica basata sulle prove.
Le evidenze suggeriscono che è proprio al livello delle giovani generazioni che si devono favorire i meccanismi di integrazione.
Nel presente contributo si analizzeranno i dati disponibili a livello nazionale e i dati raccolti nell'ambito del progetto "Contro la violenza" relativi al contesto regionale del Friuli Venezia Giulia al fine di tracciare un primo quadro dell'andamento dei fenomeni di discriminazione e intolleranza legati delle comunità straniere di prima e seconda generazione
Graph embedding procedure via dissimilarity mapping for social network comparison
No full textPaper presented to the 16th Young Statisticians Meeting. 14-16 October 2011, Rijeka, Croati
Some Issues in Estimation of Extremes in the Analysis of Network Data
No full textPractical applications of extreme value modeling can be found in many fields. For instance, in network analysis the shape and the extremal behavior of random variables describing some network characteristics is of interest. In some cases it is relevant to assess how the tail of such distributions is shaped and what is the cut-off value for which a particular distribution – typically power law like – can be fitted. The shape of the degree distribution is of great importance in network dynamic since if it follows a power law then a peculiar tie-formation mechanism (known as preferential attachment) takes place in the network, whereas if the tail is exponential a simpler random generating mechanism occurs.
Applications of extreme value theory can be found in network analysis literature. Here the common practice is to fit a power law distribution for the empirical degree distribution using the method proposed by Clauset et al. (2009). This practice is quite different with respect to the usual methods proposed in the statistical literature (Coles, 2001), especially when considering the threshold selection. In fact, the most traditional methods employ graphical diagnostics to select the threshold.
In this work we aim at comparing the results which are obtained through different methods on some network case studies in order to assess sensitiveness of substantive conclusion on the method employed. In particular we plan to compare different distributional assumption, Generalized Pareto Distribution versus Power Law versus Generalized Extreme Value distribution, and different techniques for choosing the fraction of observations to be used to estimate tail properties, graphical methods versus automatic procedures, especially the one proposed by Clauset et al. (2009).
We find out that employing graphical methods often leads to very different results with respect the automatic threshold choice of Clauset et al. (2009)
Density‐based clustering of social networks
Full text linkThe idea of the modal formulation of density-based clustering is to associate groups with the regions around the modes of the probability density function underlying the data. The correspondence between clusters and dense regions in the sample space is here exploited to discuss an extension of this approach to the analysis of social networks. Conceptually, the notion of high-density cluster fits well the one of community in a network, regarded to as a collection of individuals with dense local ties in its neighbourhood. The lack of a probabilistic notion of density in networks is turned into a strength of the proposed method, where node-wise measures that quantify the role of actors are used to derive different community configurations. The approach allows for the identification of a hierarchical structure of clusters, which may catch different degrees of resolution of the clustering structure. This feature well fits the nature of social networks, disentangling different involvements of individuals in aggregations
A comprehensive framework for solution space exploration in community detection
Full text linkCommunity detection algorithms are essential tools for understanding complex networks, yet their results often vary between runs and are affected by node input order and the presence of outliers, undermining reproducibility and interpretation. This paper addresses these issues by introducing a framework for systematic exploration of the solution space, obtained through repeated runs of a given algorithm with permuted node orders. A Bayesian model assesses convergence, estimates solution probabilities, and provides a defensible stopping rule that balances accuracy and computational cost. Building on this process, we propose a taxonomy of solution spaces that offers clear diagnostics of partition reliability across algorithms and a shared vocabulary for interpretation. Applied to a real-world network, the approach shows that different algorithms produce various types of solution space, highlighting the importance of systematic exploration of the solutions before drawing scientific conclusions
Different network typologies in patenting activity of academic inventors through time: the case of Italian chemists in the period 2000–2011
Full text linkIn this paper, we present the results of a network analysis applied to academic patent data in a subsector of the chemical field in Italy in the period 2000–2011. In particular, we analyse the micro-level interactions to point out the different network structures shaped by university-owned and university-invented patents. We detected three subnetwork typologies (labelled type A, B, and C) that exemplify different qualitative relational structures as well as different attributions to propriety rights. Type A (open science) exemplifies the typical owned patent; type B (multiple ties) represents the hybrid structure with multiple ties and involvement of academics as individuals and of universities as organisation; type C (disconnected subnetworks) represents the typical invented patent with no role of universities as organisation. The whole network seems to show a breaking point in terms of connectivity around 2005, a year that marks a change in policy rule and strategic orientation of Italian universities towards patenting. After 2005, the number of actors grew disproportionally and the network appears disconnected in several comparable components. Also, the composition in terms of subnetwork types changed. The overall picture seems to underline a big structural change dominated by the important increase of academic patenting both direct (university ownership) and indirect (increasing academic patenting)
- …
