Challenges in Social Network Research

Deriving networks and communities from individual and group attributes is an important task in understanding social groups and relations. In this work we propose a novel methodology to derive networks and communities from socio-cultural data. Our methodology is based on socio-cultural cognitive mapping (SCM) and k-NN network modularity maximization (SCM + k-NN) that produces both a latent network and community assignments of entities based upon their socio-cultural and behavioral attributes. We apply this methodology to two real-world data sets and compare the community assignments by our methodology to those communities found by k-Means, Gaussian Mixture Models, and Affinity Propagation. We then analyze the latent networks that are created by SCM + k-NN to derive novel insight into the nature of the communities. The community assignments found by SCM + k-NN are comparable to those produced by current unsupervised machine learning techniques. Additionally, in contrast to current unsupervised machine learning techniques SCM + k-NN also produces a latent network that gives additional insight into community relationships.

We propose a new approach based on bootstrapping to compare complex networks. This is an important task when we wish to compare the effect of a (policy) shock on the structure of a network. The bootstrap test compares two values of the Gini index, and the test is performed on the difference between them. The application is based on the interlocking directorship network. At the director level, Italian corporate governance is characterized by the widespread occurrence of interlocking directorates. Article 36 of Law 214/2011 prohibited interlocking directorates in the financial sector. We compare the interlocking directorship networks in 2009 (before the reform) with 2012 (after the reform) and find evidence of an asymmetric effect of the reform on the network centrality of the different companies but no significant effects on Gini indices.

This chapter investigates the latent structure of bipartite networks via a model-based clustering approach which is able to capture both latent groups of sending nodes and latent variability of the propensity of sending nodes to create links with receiving nodes within each group. This modelling approach is very flexible and can be estimated by using fast inferential approaches such as variational inference. We apply this model to the analysis of a terrorist network in order to identify the main latent groups of terrorists and their latent trait scores based on their attendance to some events.

Empirical literature shows that networks occupy an important place in a variety of economic phenomena. The first segment of related research considers the network as a date and then studies its phenomena, while the second deals with how and why networks are formed. The current study falls into this second vein and proposes a network formation model based on DEA. This work exploits the DEA methodology as a process generating relational data. In other words, the relational variable, which is the empirical basis of a network, is generated within a DEA model. In this work, a strategic economic model is developed in which the usefulness of the agents depend exclusively on direct links and their formation depends unilaterally on the decision of a single agent within the dyad. From the statistical point of view, we will propose an independent dyad model in which each dyad has a dichotomous state. We will then go on to define equilibrium as couple stability for the DEA-based network. Finally, we will conduct a micro and macro analysis of the network, estimate an ERGM based on our proposed statistical model, and compare the properties of some standard statistical models with those of our DEA-based network based on simulations.

Social interaction is mediated by computer processes at an ever-increasing rate not least because more and more people have smartphones as their everyday and habitual companions. This enables collection of a vast amount of data containing an unprecedented richness of metadata of interaction and communication. Such context information contains valuable insights for social network research and allows for qualitative grading of network structure and consecutive structural analysis. Due to the complex, heterogeneous, dynamic, and uncertain nature of such information it is yet to be considered for network analysis tasks in its entirety. In this paper, we emphasize how network analysis benefits from considering context information and identify the key challenges that have to be tackled. From an algorithmic perspective, such challenges appear on all steps of a network analysis workflow: Dynamics and uncertainty of information affects modeling networks, calculation of general metrics, calculation of centrality rankings, graph clustering, and visualization. Ultimately, novel algorithms have to be designed to combine context data and structural information to enable future context-aware network research.

During the last decades, agriculture has focused on developing more sustainable forms of land use while promoting food productivity. To face these challenges, conservation agriculture (CA), which is based on minimal soil disturbance, mulching of crop residues, and crop rotation, has been promoted as an ecosystem approach to sustainable agriculture. Using social network analysis (SNA) methods, we analyzed the adoption patterns of CA practices among 222 maize smallholder farmers in the Mexican state of Chiapas. Our findings suggest that, in the process of adopting CA practices, farmers make interrelated decisions based on network and individual attributes, rather than accepting top-down transfers, as usually promoted by institutions. The role of extension agents and other farmers is crucial for innovation dissemination, given that farmers learn different practices from different sources. This points to the need to strengthen participatory methods and promote sustainability in agriculture, rather than apply the usual hierarchical mechanisms in the innovation dissemination process.

The Naples associative dimension “from below” is considered to be a unicum at national and international level. This paper develops an exploratory research project looking at urban community hubs as social and political networks detectable by virtue of their organizational informality, exploiting the potential of data mining from social-media platforms in a user generated way. After a preliminary investigation to collect a list of active organizations in Naples, we have identified 32 organizations that have preserved some key elements over time: denomination, location, main actors. Then, we have verified and mapped the correspondence of their performed activities with the priority themes of the cultural production system defined by the European Urban Agenda. Therefore, we have built a dataset from Facebook events promoted by urban organizations and its participants; then, we have analyzed it through qualitative and quantitative data mining, and finally, we have visualized it and then proposed a cultural interpretation. The main expected results have been: firstly, to explore the proximity of the considered entities by comparison of the geographical and the social network space; secondly, to contribute to the scientific debate about the “relational ecosystem in the cities” both in theory and methods. The potential implication of this study could be considered in both fields of the interdisciplinarity of academic theories and methods, because of the proposal and use of both social sciences and statistical knowledge, and in the field of the policy strategy, because of the possibility of exploitation as a tool able to intercept the social capital of the territory and monitor the action of informal organizations operating in that territory.

This paper explores the social dynamics and the interactions between young poets and the Italian poetic network. More specifically, 70 poets were questioned about their participation in poetic activities and their cultural path. Using Social Network Analysis approach, we obtained preliminary information about the roles and structure of the Italian poetic network. Moreover, the explorative multidimensional analysis of a two-mode network was conducted using Multiple Correspondence Analysis in order to focus on the nature of relational data and the presence of asymmetry between actors and events in the network. The main results show low density network with many isolated experiences, where traditional publishers are still determinant in the building of personal career path and in the distribution network’s support. Moreover, in a market that tends to “squeeze out,” platforms and online journals are strongly emerging as places of co-production and co-distribution of the young poetry in Italy.

Policymakers wishing to enhance innovation processes in small and medium-sized enterprises increasingly channel their interventions through innovation intermediaries. However, limited empirical research exists regarding the activities and performance of intermediaries, with most contributions taking a qualitative approach and focusing on the role of intermediaries as brokers. In this paper, we analyse the extent to which innovation intermediaries, through their engagement in different activities, support the creation of communities of other agents. We use multilayer network analysis techniques to simultaneously represent the many types of interactions promoted by intermediaries. Furthermore, by originally applying the Infomap algorithm to our multilayer network, we assess the contribution of the agents involved in different activities promoted by intermediaries, and we identify the emerging multilayer communities and the intercohesive agents that span across several communities. Our analysis highlights the potential and the critical features of multilayer analysis for policy design and evaluation.

Social Network Analysis is a useful technique for studying emergent behaviours of cooperation, intervention and governance in inter-organizational networks. In this work, an empirical study of two networks of organizations operating in local territories in Southern Italy and focusing on Third Sector and welfare activities is presented. The actors are committed to experimenting a model of coordinated intervention induced by two corresponding egos which are local Caritas centres. The nodes of the two graphs are determined by combining ego-network and whole-network approaches. The weighted edges representing mutual knowledge and collaboration between nodes are determined through interviews with all actors of the local groups. It is shown that metric properties of the networks can be useful indicators to monitor and evaluate endogenous features, e.g. relational and structural embeddedness, and exogenous features characterized by homophilic mechanisms. The analysis provides insights on the networks governance of the social interacting organizations and reliable descriptors of the social processes that govern their functioning.