Computer Science

Over the past few years, there is a heated debate and serious public concerns regarding online content moderation, censorship, and the principle of free speech on the Web. To ease these concerns, social media platforms like Twitter and Facebook refined their content moderation systems to support soft moderation interventions. Soft moderation interventions refer to warning labels attached to potentially questionable or harmful content to inform other users about the content and its nature while the content remains accessible, hence alleviating concerns related to censorship and free speech. In this work, we perform one of the first empirical studies on soft moderation interventions on Twitter. Using a mixed-methods approach, we study the users who share tweets with warning labels on Twitter and their political leaning, the engagement that these tweets receive, and how users interact with tweets that have warning labels. Among other things, we find that 72% of the tweets with warning labels are shared by Republicans, while only 11% are shared by Democrats. By analyzing content engagement, we find that tweets with warning labels had more engagement compared to tweets without warning labels. Also, we qualitatively analyze how users interact with content that has warning labels finding that the most popular interactions are related to further debunking false claims, mocking the author or content of the disputed tweet, and further reinforcing or resharing false claims. Finally, we describe concrete examples of inconsistencies, such as warning labels that are incorrectly added or warning labels that are not added on tweets despite sharing questionable and potentially harmful information.

WhatsApp is the most popular messaging app in the world. Due to its popularity, WhatsApp has become a powerful and cheap tool for political campaigning being widely used during the 2019 Indian general election, where it was used to connect to the voters on a large scale. Along with the campaigning, there have been reports that WhatsApp has also become a breeding ground for harmful speech against various protected groups and religious minorities. Many such messages attempt to instil fear among the population about a specific (minority) community. According to research on inter-group conflict, such `fear speech' messages could have a lasting impact and might lead to real offline violence. In this paper, we perform the first large scale study on fear speech across thousands of public WhatsApp groups discussing politics in India. We curate a new dataset and try to characterize fear speech from this dataset. We observe that users writing fear speech messages use various events and symbols to create the illusion of fear among the reader about a target community. We build models to classify fear speech and observe that current state-of-the-art NLP models do not perform well at this task. Fear speech messages tend to spread faster and could potentially go undetected by classifiers built to detect traditional toxic speech due to their low toxic nature. Finally, using a novel methodology to target users with Facebook ads, we conduct a survey among the users of these WhatsApp groups to understand the types of users who consume and share fear speech. We believe that this work opens up new research questions that are very different from tackling hate speech which the research community has been traditionally involved in.

Against the backdrop of a social media reckoning, this paper seeks to demonstrate the potential of social tools to build virtuous behaviours online. We must assume that human behaviour is flawed, the truth can be elusive, and as communities we must commit to mechanisms to encourage virtuous social digital behaviours. Societies that use social platforms should be inclusive, responsive to evidence, limit punitive actions and allow productive discord and respectful disagreement. Social media success, we argue, is in the hypothesis. Documents are valuable to the degree that they are evidence in service of, or to challenge an idea for a purpose. We outline how a Bayesian social platform can facilitate virtuous behaviours to build evidence-based collective rationality. The chapter outlines the epistemic architecture of the platform's algorithms and user interface in conjunction with explicit community management to ensure psychological safety. The BetterBeliefs platform rewards users who demonstrate epistemically virtuous behaviours and exports evidence-based propositions for decision-making. A Bayesian social network can make virtuous ideas powerful.

Thanks to platforms such as Twitter and Facebook, people can know facts and events that otherwise would have been silenced. However, social media significantly contribute also to fast spreading biased and false news while targeting specific segments of the population. We have seen how false information can be spread using automated accounts, known as bots. Using Twitter as a benchmark, we investigate behavioural attitudes of so called `credulous' users, i.e., genuine accounts following many bots. Leveraging our previous work, where supervised learning is successfully applied to single out credulous users, we improve the classification task with a detailed features' analysis and provide evidence that simple and lightweight features are crucial to detect such users. Furthermore, we study the differences in the way credulous and not credulous users interact with bots and discover that credulous users tend to amplify more the content posted by bots and argue that their detection can be instrumental to get useful information on possible dissemination of spam content, propaganda, and, in general, little or no reliable information.

People's opinions evolve over time as they interact with their friends, family, colleagues, and others. In the study of opinion dynamics on networks, one often encodes interactions between people in the form of dyadic relationships, but many social interactions in real life are polyadic (i.e., they involve three or more people). In this paper, we extend an asynchronous bounded-confidence model (BCM) on graphs, in which nodes are connected pairwise by edges, to hypergraphs. We show that our hypergraph BCM converges to consensus under a wide range of initial conditions for the opinions of the nodes. We show that, under suitable conditions, echo chambers can form on hypergraphs with community structure. We also observe that the opinions of individuals can sometimes jump from one opinion cluster to another in a single time step, a phenomenon (which we call "opinion jumping") that is not possible in standard dyadic BCMs. We also show that there is a phase transition in the convergence time on the complete hypergraph when the variance ? 2 of the initial opinion distribution equals the confidence bound c . Therefore, to determine the convergence properties of our hypergraph BCM when the variance and the number of hyperedges are both large, it is necessary to use analytical methods instead of relying only on Monte Carlo simulations.

Given that Proof-of-Work (PoW) and Proof-of-Stake (PoS) are plutocratic, and other common consensus protocols are mostly permission-based, we look for a consensus protocol that will suit the needs of e-Democracy. In particular, what we need is a distributed ledger that will record and, to the possible extent, execute the public will. We propose a combination of any given permission-based protocol together with a trust graph between the nodes, which supplies the required permission for new nodes. As a result, the consensus protocol reaches consensus at every iteration between a known list of agents and then updates this list between iterations. This paper is based on prior work that shows the conditions under which a community can grow while maintaining a bounded number of byzantines. It combines a permission-based consensus protocol (such as pBFT) with a community expansion algorithm (such as the one in the prior work) to arrive at a consensus protocol in which the set of agents can change in time, while being sybil-resilient.

Many governments impose traditional censorship methods on social media platforms. Instead of removing it completely, many social media companies, including Twitter, only withhold the content from the requesting country. This makes such content still accessible outside of the censored region, allowing for an excellent setting in which to study government censorship on social media. We mine such content using the Internet Archive's Twitter Stream Grab. We release a dataset of 583,437 tweets by 155,715 users that were censored between 2012-2020 July. We also release 4,301 accounts that were censored in their entirety. Additionally, we release a set of 22,083,759 supplemental tweets made up of all tweets by users with at least one censored tweet as well as instances of other users retweeting the censored user. We provide an exploratory analysis of this dataset. Our dataset will not only aid in the study of government censorship but will also aid in studying hate speech detection and the effect of censorship on social media users. The dataset is publicly available at this https URL

For e-commerce platforms such as Taobao and Amazon, advertisers play an important role in the entire digital ecosystem: their behaviors explicitly influence users' browsing and shopping experience; more importantly, advertiser's expenditure on advertising constitutes a primary source of platform revenue. Therefore, providing better services for advertisers is essential for the long-term prosperity for e-commerce platforms. To achieve this goal, the ad platform needs to have an in-depth understanding of advertisers in terms of both their marketing intents and satisfaction over the advertising performance, based on which further optimization could be carried out to service the advertisers in the correct direction. In this paper, we propose a novel Deep Satisfaction Prediction Network (DSPN), which models advertiser intent and satisfaction simultaneously. It employs a two-stage network structure where advertiser intent vector and satisfaction are jointly learned by considering the features of advertiser's action information and advertising performance indicators. Experiments on an Alibaba advertisement dataset and online evaluations show that our proposed DSPN outperforms state-of-the-art baselines and has stable performance in terms of AUC in the online environment. Further analyses show that DSPN not only predicts advertisers' satisfaction accurately but also learns an explainable advertiser intent, revealing the opportunities to optimize the advertising performance further.

All societies have been much more bipolar over the past few years, particularly after the emergence of online social networks and media. In fact, the gap between the two ends of social spectrum is going to be even deeper after the spread of new media. In this circumstance, social polarization has been a growing concern among socialists and computer science experts because of the detrimental impact which online social networks can have on societies by adding fuel to the fire of extremism. Several types of research were conducted for proposing measures to calculate the controversy level in social networks, afterward, to reduce controversy among contradicting viewpoints, for example, by exposing opinions of one side to other side's members. Most of the attempts for quantifying social networks' controversy have considered the networks in their most primary forms, without any attributes. Although these kinds of researches provide platform-free algorithms to be used in different social networks, they are not able to take into account a great deal of useful information provided by users (node attributes).To surmount this shortcoming, we propose a framework to be utilized in different networks with different attributes. We propelled some Biased Random Walks (BRW) to find their path from start point to an initially unknown endpoint with respect to initial energy of start node and energy loss of nodes on the path. We extracted structural attribute of networks, using node2vec, and compared it with state-of-the-art algorithms, and showed its accuracy. Then, we extracted some content attributes of user and analyze their effects on the results of our algorithm. BRW is compared with another state-of-the-art controversy measuring algorithm. Then, its changes in different level of controversy in Persian Twitter is considered to show how it works in different circumstance.

Cultural Ecosystem Services (CES) assessment at large scales is crucial in marine ecosystems as they reflect key physical and cognitive interactions between humans and nature. The analysis of social media data with graph theory is a promising approach to provide global information on users' perceptions for different marine ecosystems. Fourteen areas were selected to illustrate the use of graph theory on social media data. The selected areas, known to protect key recreational, educational and heritage attributes of marine ecosystems, were investigated to identify variability in users' preferences. Instagram data (i.e., hashtags associated to photos) was extracted for each area allowing an in-depth assessment of the CES most appreciated by the users. Hashtags were analysed using network centrality measures to identify clusters of words, aspects not normally captured by traditional photo content analysis. The emergent properties of networks of hashtags were explored to characterise visitors' preferences (e.g., cultural heritage or nature appreciation), activities (e.g., diving or hiking), preferred habitats and species (e.g. forest, beach, penguins), and feelings (e.g., happiness or place identity). Network analysis on Instagram hashtags allowed delineating the users' discourse around a natural area, which provides crucial information for effective management of popular natural spaces for people.