Cross-correlation patterns in social opinion formation with sequential data

In this paper, we propose a new model for continuous time opinion dynamics on an evolving network. As opposed to existing models, in which the network typically evolves by discretely adding or removing edges, we instead propose a model for opinion formation which is coupled to a network evolving through a system of ordinary differential equations for the edge weights. We interpret each edge weight as the strength of the relationship between a pair of individuals, with edges increasing in weight if pairs continually listen to each other’s opinions and decreasing if not. We investigate the impact of various edge dynamics at different timescales on the opinion dynamics itself. This is done partly through analytic results and partly through extensive numerical simulations of two case studies: one using bounded confidence interaction dynamics in the opinion formation process (as in the classical Hegselmann–Krause model) and one using an exponentially decaying interaction function. We find that the dynamic edge weights can have a significant impact on the opinion formation process, since they may result in consensus formation but can also reinforce polarisation. Overall, the proposed modelling approach allows us to quantify and investigate how the network and opinion dynamics influence each other.

The opinion dynamics literature argues that the way people perceive social influence depends not only on the opinions of interacting individuals, but also on the individuals’ non-opinion characteristics, such as age, education, gender, or place of residence. The current paper advances this line of research by studying longitudinal data that describe the opinion dynamics of a large sample (~30,000) of online social network users, all citizens of one city. Using these data, we systematically investigate the effects of users’ demographic (age, gender) and structural (degree centrality, the number of common friends) properties on opinion formation processes. We revealed that females are less easily influenced than males. Next, we found that individuals that are characterized by similar ages have more chances to reach a consensus. Additionally, we report that individuals who have many common peers find an agreement more often. We also demonstrated that the impacts of these effects are virtually the same, and despite being statistically significant, are far less strong than that of opinion-related features: knowing the current opinion of an individual and, what is even more important, the distance in opinions between this individual and the person that attempts to influence the individual is much more valuable. Next, after conducting a series of simulations with an agent-based model, we revealed that accounting for non-opinion characteristics may lead to not very sound but statistically significant changes in the macroscopic predictions of the populations of opinion camps, primarily among the agents with radical opinions (≈3% of all votes). In turn, predictions for the populations of neutral individuals are virtually the same. In addition, we demonstrated that the accumulative effect of non-opinion features on opinion dynamics is seriously moderated by whether the underlying social network correlates with the agents’ characteristics. After applying the procedure of random shuffling (in which the agents and their characteristics were randomly scattered over the network), the macroscopic predictions have changed by ≈9% of all votes. What is interesting is that the population of neutral agents was again not affected by this intervention.

Rational forecasts or social opinion dynamics? Identification of interaction effects in a business climate survey

Games are fundamentally a social activity. The effects of this foundation can be felt at every level - from the social negotiation of rules, through cooperation and collaboration between players during the game, to the effects of relationships and social status on play. Social effects can change the way the game is played, but the mechanics of games can also affect the patterns of social behaviours of the players. The arrangement of game mechanics and interfaces together defines a ``social architecture. This architecture is not limited to directly social mechanics such as trading and messaging - the game design itself has a holistic effect on social activity. This dissertation frames games around these social aspects, and focuses on analysis of the patterns that emerge from these playful interactions. Firstly, a model is defined to understand games based on the social effects of play, and these effects explored based on the varying impact they have on the play experience. Mischief and deviance is also investigated as forces that challenge these social effects in and around games. Based on interaction data gathered from server logs of experimental social games, social network analysis is used as a tool to uncover the macroscopic social architectures formed by each design. This allows the use of quantitative methods to understand the nature of the relationship between game design and the social patterns that emerge around games in play.Key findings confirm that social activity follows a heavy-tailed distribution - a small number of ``hardcore players are responsible for a disproportionately large number of interactions in the community of the game. Further than this, the connections between active hardcore and the rest of the player base show that without the hardcore users, the community of games as ``small worlds would collapse, with large numbers of players being separated from the society within a game. The emergence of grouping behaviour is investigated based on the effect of social feedback. Following findings of social psychology in non-game environments, evidence is provided that highlights the effect of socio-contextual feedback on players forming strongly bound tribal groups within games. The communities formed through the play of games can be described in terms of network graphs - webs of interactions flowing around a network of players. Social network analyses of social games show the emergence of patterns of reciprocity, clustering and tribal behaviours among the players. The evidence also shows that the collections of game mechanics, or social architectures, of games have a predictable effect on the wider social patterns of the players. As such, this suggests games can be specifically engineered for social effects based on changes in the patterns of interactions, and issues around mechanical or interface elements can be identified based on anomalies observed in the network graph of player interactions.Together, this dissertation provides a link between the theoretical ideas around social play to the measurable effects of social behaviours of players within games. It proves that game designs, as mechanical systems, have a demonstrable effect on the social patterns of play, and that these patterns can be examined and used to engineer better game designs for the benefit of social experience.

Abstractsocial networks Online provide a globally available, large-scale infrastructure for people to exchange information and ideas. A topic of great interest in internet research is how to model this information exchange and, in particular, how to model and analyze the effects of interpersonal influence on processes such as information diffusion, influence propagation, and opinion formation. Recent empirical studies indicate that, in order to accurately model communication in online social networks, it is important to consider not just relationships between individuals, but also the frequency with which these individuals interact. We study a model of opinion formation in social networks proposed by De Groot and Lehrer and show how this model can be extended to include interaction frequency. We prove that, for the purposes of analysis and design, the opinion formation process with probabilistic interactions can be accurately approximated by a deterministic system where edge weights are adjusted for the probability of interaction. We also present simulations that illustrate the effects of different interaction frequencies on the opinion dynamics using real-world social network graphs.KeywordsOpinion DynamicsSocial InfluenceComplex NetworkAgent-based Modeling

The referendums on the European Constitutional Treaty which took place in Spain, France, the Netherlands and Luxembourg in 2005 provided scholars with a unique opportunity to examine more closely the process by which voters form and change opinions on complex and unfamiliar political questions. Only a few voters in these instances came to the referendums on the European Constitution with strongly held and fully formed opinions regarding the issue(s) involved. This is largely because constitutional matters tend, under most circumstances, to be the exclusive preserve of political elites. When they are introduced into the realm of mass electoral politics, voters must begin a complex learning process in order to become familiar with the underlying arguments and issues. We know from the study of referendums at other times and in other places that this process of opinion formation is complex and highly uncertain, depending on the issues involved (Zaller, 1991, 1992; LeDuc and Svensson, 2002). According to a special Eurobarometer published in January 2005, only 11 per cent of respondents indicated that they knew much about the contents of the Constitutional Treaty.1 About a third of those surveyed reported that they had not heard of the treaty at all, while the balance (56 per cent) said that they had heard something about it but knew very little about its substance. Referendums on issues such as these with which the mass public is largely unfamiliar often tend to be highly volatile and therefore unpredictable events.

This paper defines two different ways of the process of the opinion formation and focuses on the effect of the proportion of the two strategies of process and the structure of the network. A multiopinion model is proposed in this paper, which includes two strategies of opinion formation. At the first part, the change of the structured network and the change of the single node are used as two strategies of the process of the opinion formation. We focus on how the proportion of the two strategies can affect the outcome of the process. At the second part, as the proportion of the two strategies is fixed, the edges are dense in the community and sparse outside. Thus we can construct a bifurcation diagram to be verified through experimental study. The phase transition is studied in the network which contains more than four opinions and two strategies of process. Our results show that the size of the group and the dense of edges are the two important features for the process of opinion formation.

Mass participation of the members of a society in discussions to resolve issues related to a topic leads to forming public opinion. The timeline of the underlying dynamics goes through several distinguishable phases, and experiences transition from one to another. After initiated by concerned individuals, it draws active attention from almost everyone, and with time progression, people's participation starts declining as the issues are resolved or lost attraction. The existing works in the literature to capture the opinion formation process pay attention to model the dynamics in its active phase and thus ignore the other phases and the corresponding phase transitions. Trust relationships among the participants dynamically shape their interactions in different stages of the dynamics. Existing works fail to incorporate trust in defining the extent of influence one has on others, as they define the social relationships in the opinion space. To address this issue, we adopt simulated annealing to model the transitional behaviour of the dynamics, and then, amalgamate peoples relationships in the trust space with that in the opinion space to define the meta-heuristics of the algorithm for capturing the dynamical properties of the process. Finally, through simulation, we observe that our model is insightful in representing peoples' evolving behaviour in the different stages of opinion formation process, and consequently, can capture the various properties of the steady-state outcomes of the dynamics.

On a discrete generalized kinetic approach for modelling persuader’s influence in opinion formation processes

Social effects represent the psychological (emotional, cognitive, and motivational) reactions evoked in other people by the expression of traits in behavior and emotion. From the transactional view on personality, studying the psycholexical structures of social effects can help to discover unique vs. common thought and behavior patterns, affects, and motivations, which are primarily related to personality dispositions. Thus, we developed the comprehensive taxonomy of social effects following the principles of the psycholexical approach. In the first study, two judges selected 9,625 person-descriptive terms—adjectives, type-nouns, attribute-nouns, and participles—from the Dictionary of the Standard Lithuanian Language. In the second study, six judges classified all the selected descriptors using German psycholexical methodology. Finally, a principal component analysis was performed, followed by varimax rotation for the 208 social-effect descriptors, separately for ipsatized self-ratings and observer-ratings from 203 to 204 Lithuanian students, respectively. We found out that the five-component solution was the best fit for self-ratings, whereas for observer-ratings it was a four-component structure. In this article, we present the results from the factor analyses and discuss our findings in the context of previous studies, as well as cross-language personality models.

Information Inundation and Polarization In the paper “Information Inundation on Platforms and Implications,” Gad Allon, Kimon Drakopoulos, and Vahideh Manshadi study the process of learning and opinion formation in the presence of platforms. The paper is motivated by the relatively recent changes in how people access and consume news. Specifically, more and more people access news through social platforms, such as Facebook, Reddit, or Twitter. The paper tries to answer a simple question: why is it that we become more polarized even though we have more information sources than ever? In this paper, the authors assume that people are almost rational and are not easily biased as a result of psychological biases. Yet the authors show that, even if people are good “statisticians” in the sense that they try to find the information that reduces their ignorance by the most significant amount, they choose to consume information that slows their ability to learn. In fact, the authors show that screening information sources results in polarization: everyone is sort of “stuck” in their side of the political map. We are essentially illustrating a mechanism that generates confirmation bias even for a seemingly rational reader.

EFFECT OF LEADER'S STRATEGY ON OPINION FORMATION IN NETWORKED SOCIETIES WITH LOCAL INTERACTIONS

Unlike the internalizing domain of psychological problems, externalizing problems are not defined by the experience of inherently distressing emotions such as anxiety or sadness. Rather, they are defined by patterns of behavior that interfere with living a successful and satisfying life. People with high levels of externalizing problems are often very distressed, but they are mostly distressed by the consequences of their actions. Their inattentive, hyperactive, impulsive, irritable, oppositional, antisocial, callous, psychopathic, narcissistic, and histrionic behaviors and their abuse of, and dependence on, psychoactive substances interfere with functioning in school, work, and other settings; alienate peers and family members; and can lead to serious social consequences such as family conflict, being fired from employment, or incarceration.

The advent of Generative Adversarial Network (GAN) architectures has given anyone the ability of generating incredibly realistic synthetic imagery. The malicious diffusion of GAN-generated images may lead to serious social and political consequences (e.g., fake news spreading, opinion formation, etc.). It is therefore important to regulate the widespread distribution of synthetic imagery by developing solutions able to detect them. In this paper, we study the possibility of using Benford’s law to discriminate GAN-generated images from natural photographs. Benford’s law describes the distribution of the most significant digit for quantized Discrete Cosine Transform (DCT) coefficients. Extending and generalizing this property, we show that it is possible to extract a compact feature vector from an image. This feature vector can be fed to an extremely simple classifier for GAN-generated image detection purpose.

Understanding and measuring sentence acceptability is of fundamental importance for linguists, but although many measures for doing so have been developed, relatively little is known about some of their psychometric properties. In this paper we evaluate within- and between-participant test-retest reliability on a wide range of measures of sentence acceptability. Doing so allows us to estimate how much of the variability within each measure is due to factors including participant-level individual differences, sample size, response styles, and item effects. The measures examined include Likert scales, two versions of forced-choice judgments, magnitude estimation, and a novel measure based on Thurstonian approaches in psychophysics. We reproduce previous findings of high between-participant reliability within and across measures, and extend these results to a generally high reliability within individual items and individual people. Our results indicate that Likert scales and the Thurstonian approach produce the most stable and reliable acceptability measures and do so with smaller sample sizes than the other measures. Moreover, their agreement with each other suggests that the limitation of a discrete Likert scale does not impose a significant degree of structure on the resulting acceptability judgments.