Social networks, forums, and social media have emerged as global platforms for forming and shaping opinions on a broad spectrum of topics like politics, sports and entertainment. Users (also called actors) often update their evolving opinions, influenced through discussions with other users. Theoretical models and their analysis on understanding opinion dynamics in social networks abound in the literature. However, these models are often based on concepts from statistical physics. Their goal is to establish various regulatory phenomena like steady-state consensus or bifurcation. Analysis of transient effects is largely avoided. Moreover, many of these studies assume that actors opinions are observed globally and synchronously, which is rarely realistic. In this paper, we initiate an investigation into a family of novel data-driven influence models that accurately learn and fit realistic observations. We estimate and do not presume edge strengths from observed opinions at nodes. Our influence models are linear, but not necessarily positive or row stochastic in nature. As a consequence, unlike the previous studies, they do not depend on system stability or convergence during the observation period. Furthermore, our models take into account a wide variety of data collection scenarios. In particular, they are robust to missing observations for several time steps after an actor has changed its opinion. In addition, we consider scenarios where opinion observations may be available only for aggregated clusters of nodes a practical restriction often imposed to ensure privacy. Finally, to provide a conceptually interpretable design of edge influence, we offer a relatively frugal variant of our influence model, where the strength of influence between two connecting nodes depend on the node attributes (demography, personality, expertise etc.). Such an approach reduces the number of model parameters, reduces overfitting, and offers a tractable and explicable sketch of edge-influences in the context of opinion dynamics. With six real-life datasets crawled from Twitter and Reddit, as well as three more datasets collected from in-house experiments (with 102 volunteers), our proposed system gives significant accuracy boost over four state-of-the-art baselines. We also observe that a careful design of edge strengths using node properties is crucial, since it offers substantially better performance than the one with independent edge weights.
Cyberbullying and cyberaggression are increasingly worrisome phenomena that affect people across all demographics. Already in 2014, more than half of young social media users worldwide experienced them in some form, being exposed to prolonged and/or coordinated digital harassment. Victims can experience a wide range of emotional consequences such as embarrassment, depression, isolation from other community members, which can lead to even more serious consequences such as suicide attempts. Nevertheless, tools and technologies to understand and mitigate it are scarce and mostly ineffective. In this paper, we take the first concrete steps to understand the characteristics of abusive behavior in Twitter, one of today's largest social networks. We analyze 1.2 million users and 2 million tweets, comparing users participating in discussions around seemingly normal topics like the NBA, to those more likely to be hate-related, such as the Gamergate controversy or the gender pay inequality at the BBC. We also explore specific manifestations of abusive behavior, i.e., cyberbullying and cyberaggression, in one of the hate-related communities (Gamergate). We present a robust methodology to distinguish bullies and aggressors from regular users by considering text, user, and network based attributes. Using various state-of-the-art machine learning algorithms, we can classify these accounts with over 90% accuracy and AUC. Finally, we look at the current status of the Twitter accounts of users marked as abusive by our methodology and discuss the performance of the mechanisms used by Twitter to suspend users.