Session F28: Physical Approaches to Social Modeling

Online Networks and the Diffusion of Protests

Undoubtedly, online social networks have an enormous impact on opinions and cultural trends. Also, these platforms have revealed as a fundamental organizing mechanism in country-wide social movements. Recent events in the Middle East and North Africa (the wave of protests in the Arab world), across Europe (in the form of anti-cuts demonstrations or riots) and United States (the OWS movement) have generated much discussion on how digital media is connected to the diffusion of protests. In this talk, we investigate the mechanisms driving the emergence, development and stabilization of unrest movements in Spain and the USA by analyzing data from Twitter. Messages related to the protests are analyzed at both static and dynamic levels. We show that the online trace of the protests provides a unique opportunity to tackle central issues like recruitment patterns, information cascades and their spatiotemporal dynamics. Our findings shed light on the connection between online networks and social movements, and offer an empirical test to elusive sociological questions about collective action.   

Role of Committed Minorities in Times of Crisis

The surprising social phenomena of the Arab Spring and the Occupy Wall Street movement posit the question of whether the active role of committed groups may produce political changes of significant importance. Under what conditions are the convictions of a minority going to dominate the future direction of a society? We use a Cooperative Decision Making (CDM) model to study the effect of committed minorities on group behavior in time of crisis. The CDM model has been shown to generate consensus through a phase-transition process that at criticality establishes long-range correlations among the individuals within a model society. In a condition of high consensus, the correlation function vanishes, thereby making the network recover the ordinary locality condition. However, this state is not permanent and times of crisis occur when there is an ambiguity concerning a given social issue. The correlation function within the cooperative system becomes similarly extended as it is observed at criticality. This combination of independence (free will) and long-range correlation makes it possible for very small but committed minorities to produce substantial changes in social consensus.   

Beating the news using social media: the case study of American Idol

We present a contribution to the debate on the predictability of social events using big data analytics. We focus on the elimination of contestants in the American Idol TV shows as an example of a well defined electoral phenomenon to assess the predictive power of twitter signals. We provide evidence that Twitter activity during the time span defined by the TV show airing and the voting period following it allows the anticipation of the voting outcome. Twitter data have been analyzed to attempt the winner prediction ahead of the airing of the official result. We also show that the fraction of Tweets that contain geolocation information allows us to map the fanbase of each contestant, both within the US and abroad, showing that strong regional polarizations occur. The geolocalized data are crucial for the correct prediction of the final outcome of the show, pointing out the importance of considering information beyond the aggregated twitter signal. Although American Idol voting is just a minimal and simplified version of complex societal phenomena, this work shows that the volume of information available in online systems permits the real time gathering of quantitative indicators that may be able to anticipate the future unfolding of opinion formation events.   

Heterogeneity of Human Activity Levels Gives Rise to Power-Law Distribution in Online Social Networks

It is well established that the distribution of social ties (degree) of an individual in a social network follows a power-law. How this heavy-tailed distribution arises in practice, however, has not been conclusively demonstrated. Mechanisms of ``preferential-attachment'' and optimization are often cited as the origin of heavy-tailed degree distributions. Our data indicate that there is a different cause for these phenomena. For different social networks we find an intrinsic relationship degree and activity (number of posts, edits etc): The degree distribution is entirely random except for its mean value which depends deterministically on the volume of the users' activity. This suggests that heavy-tailed degree distribution is a consequence of the intrinsic activity of users. More importantly, human activity deterministically affects the mean success at establishing links in a social network, and the specific degree of a given user is otherwise random following a ``maximum entropy attachment'' model.   

Nonlinear Opinion Dynamics on Networks

A model which treats group decision making as nonlinear opinion dynamics occurring over a network is presented. The model makes predictions regarding the interaction of network structure and initial disagreement level upon decision outcomes and consensus formation. The model displays bifurcations at high disagreement levels which lead to behaviors that are qualitatively distinct from those at low disagreement. For example, at high disagreement, the model exhibits asymmetric, majority rule outcomes that arise even when the system is symmetric with respect to the distribution of initial opinions and network structure. Analytical approximations for the bifurcation boundaries agree well with numerically-determined boundaries. An ongoing experimental effort involving the use of online discussion groups to test the model predictions is briefly described.   

Language Geography from Microblogging Platforms

Microblogging platforms have now become major open source indicators for complex social interactions. With the advent of smartphones, the everincreasing mobile Internet traffic gives us the unprecedented opportunity to complement studies of complex social phenomena with real-time location information. In this work, we show that the data nowadays accessible allows for detailed studies at different scales, ranging from country-level aggregate analysis to the analysis of linguistic communities withing specific neighborhoods. The high resolution and coverage of this data permits us to investigate such issues as the linguistic homogeneity of different countries, touristic seasonal patterns within countries, and the geographical distribution of different languages in bilingual regions. This work highlights the potentialities of geolocalized studies of open data sources that can provide an extremely detailed picture of the language geography.   

The rise and fall of social communities: Cascades of followers triggered by innovators

New scientific ideas as well as key political messages, consumer products, advertisement strategies and art trends are originally adopted by a small number of pioneers who innovate and develop the ``new ideas''. When these innovators migrate to develop the novel idea, their former social network gradually weakens its grips as followers migrate too. As a result, an internal ``cascade of followers'' starts immediately thereafter speeding up the extinction of the entire original network. A fundamental problem in network theory is to determine the minimum number of pioneers that, upon leaving, will disintegrate their social network. Here, we first employ empirical analyses of collaboration networks of scientists to show that these communities are extremely fragile with regard to the departure of a few pioneers. This process can be mapped out on a percolation model in a correlated graph crucially augmented with outgoing ``influence links''. Analytical solutions predict phase transitions, either abrupt or continuous, where networks are disintegrated through cascades of followers as in the empirical data. The theory provides a framework to predict the vulnerability of a large class of networks containing influence links ranging from social and infrastructure networks to financial systems and markets.   

Extraordinary variability and sharp transitions in a maximally frustrated dynamic network

Most previous studies of complex networks have focused on single, static networks. However, in the real world, networks are dynamic and interconnected. Inspired by the presence of extroverts and introverts in the general population, we investigate a highly simplified model of a social network, involving two types of nodes: one preferring the highest degree possible, and one preferring no connections whatsoever. There are only two control parameters in the model: the number of ``introvert'' and ``extrovert'' nodes, $N_I$ and $N_E$. Our key findings are as follows: As a function of $N_I$ and $N_E$, the system exhibits a highly unusual transition, displaying extraordinary fluctuations (as in 2nd order transitions) and discontinuous jumps (characteristic of 1st order transitions). Most remarkably, the system can be described by an Ising-like Hamiltonian with long-range multi-spin interactions and some of its properties can be obtained analytically. This is in stark contrast with other dynamic network models which rely almost exclusively on simulations.   

Self-organization of plutonomy in competitive societies

The plutonomy, where the top 1 percent of households accounts for more wealth than the bottom 99 percent, is an extreme form of hierarchy. It is an important question to find out what traits of competitive society lead to the extreme form of hierarchy. We investigate conditions for emergence of plutonomy in two model competitive societies. In a model where individuals make random walk and fight when they meet, we show that the order of move plays an important role and the plutonomy emerges when individuals have the same right to move first. For a model society where individuals participate in a competition with equal right, we show that the plutonomy can be self-organized when individuals grouped into several classes compete with those in the same class for a certain period (season) and they are regrouped at the end of every season. We also discuss various features of the emergence of plutonomy.   

Popularity-Driven Networking

We investigate the growth of connectivity in a network. In our model, starting with a set of disjoint nodes, links are added sequentially. Each link connects two nodes, and the connection rate governing this random process is proportional to the degrees of the two nodes. Interestingly, this network exhibits two abrupt transitions, both occurring at finite times. The first is a percolation transition in which a giant component, containing a finite fraction of all nodes, is born. The second is a condensation transition in which the entire system condenses into a single, fully connected, component. We derive the size distribution of connected components as well as the degree distribution, which is purely exponential throughout the evolution. Furthermore, we present a criterion for the emergence of sudden condensation for general homogeneous connection rates.   

Threshold model with multiple initiators

In social networks, adoption of a new behavior or opinion by an agent strongly depends on its neighborhood. We study the threshold model where every node is in one of two possible states (0 or 1) and a node in state '0' changes to '1' if at least a threshold fraction $\phi$ of its neighbors are already in state '1'[M. Granovetter, AJS, Vol. 83, No. 6]. Initially all nodes are in state '0' except initiators. Previous studies have shown that a small seed of such initiators can give rise to large cascades if $\phi$ is less than some critical $\phi_{c}$. The focus of our work is the effect of the size of the initiator fraction $p$ on the size of the cascade for different threshold values, on empirical networks as well as stylized models of social networks. We observe that global cascades are possible for arbitrary values of $\phi$, if $p$ is sufficiently large. We find that there exists a critical $p_{c}$ (unique for every $\phi$), such that $p\geq p_{c}$ results in global cascades whereas for $p [Preview Abstract]

Spreading of infectious diseases considering age contact patterns for Latin America

The dynamics of infectious diseases strongly depends on the structure of the social contact patterns among individuals. In order to have an accurate estimate of the impact of epidemic outbreaks and which effective control measures to take, we need an appropriate description of these patterns. A simple way to improve the homogeneous mixing assumption is to introduce age contact patterns. Here we follow the approach of Fumanelli et al (PLoS Computational Biology, 8(9):e1002673, 2012) to estimate the age mixing patterns of virtual populations using highly detailed census data for Argentina, Brazil and Mexico. Considering age contact matrices for these countries we study the epidemiological relevant quantities and their relation with the sociodemographic data. Our results show that even for the same country the impact of epidemics outbreaks could be very different when we consider age contact matrices. This results can be explained as a result of a change in the average age of the population in the different regions of the countries. This study also provides the first estimates of contact matrices for Latin American countries.   

S-curves and the Mechanism of Propagation in Language Change

Linguists have proposed a wide variety of mechanisms for the propagation of a linguistic innovation through the speech community. The complexity of social systems makes it difficult to evaluate the different mechanisms empirically. We introduce a four-way typology of mechanisms and provide mathematical definitions based on the symmetries that may (or may not) be present between different speakers in the community and/or between different linguistic variants. As in physics, such symmetries impose strong constraints on the patterns of change that may emerge as a result. In particular, we conclude that the widely observed empirical pattern of an S-curve temporal trajectory of change can be captured only by theories that invoke a pre-existing shared preference among speakers for the incoming variant.