A5086827245- Social Media and Politics
- Misinformation and Its Impacts
- Complex Network Analysis Techniques
- Opinion Dynamics and Social Influence
- Hate Speech and Cyberbullying Detection
- Media Influence and Politics
- Wikis in Education and Collaboration
- Open Source Software Innovations
- Spam and Phishing Detection
- Media Studies and Communication
- Impact of Technology on Adolescents
- Game Theory and Voting Systems
- Advanced Text Analysis Techniques
- Topic Modeling
- Privacy, Security, and Data Protection
- Complexity and Algorithms in Graphs
- Digital Communication and Language
- Digital Games and Media
- Knowledge Management and Sharing
- Mobile Crowdsensing and Crowdsourcing
- Consumer Market Behavior and Pricing
- Chemical Thermodynamics and Molecular Structure
- FinTech, Crowdfunding, Digital Finance
- Turkey's Politics and Society
- Multi-Agent Systems and Negotiation
University of Michigan
2016-2025
State Street (United States)
2024
Ann Arbor Center for Independent Living
2016
Microsoft Research (United Kingdom)
2013-2015
Microsoft (United States)
2015
University of California, Santa Barbara
2008-2014
In this work, we study the notion of competing campaigns in a social network and address problem influence limitation where "bad" campaign starts propagating from certain node use limiting to counteract effect misinformation. The can be summarized as identifying subset individuals that need convinced adopt (or "good") so minimize number people at end both propagation processes. We show optimization is NP-hard provide approximation guarantees for greedy solution various definitions by proving...
It is widely thought that news organizations exhibit ideological bias, but rigorously quantifying such slant has proven methodologically challenging. Through a combination of machine-learning and crowdsourcing techniques, we investigate the selection framing political issues in fifteen major US outlets. Starting with 803,146 stories published over twelve months, first used supervised learning algorithms to identify 14 percent articles pertaining events. We then recruited 749 online human...
Since December 2019, COVID-19 has been spreading rapidly across the world. Not surprisingly, conversation about is also increasing. This article a first look at amount of taking place on social media, specifically Twitter, with respect to COVID-19, themes discussion, where discussion emerging from, myths shared virus, and how much it connected other high low quality information Internet through URL links. Our preliminary findings suggest that meaningful spatio-temporal relationship exists...
Online communities about similar topics may maintain very different norms of interaction. Past research identifies many processes that contribute to maintaining stable norms, including self-selection, pre-entry learning, post-entry and retention. We analyzed political subreddits had distinctive, levels toxic comments on Reddit, in order identify the relative contribution these four processes. Surprisingly, we find largest source norm stability is learning. That is, newcomers' first...
Written by Michelle A. Amazeen, Fabrício Benevenuto, Nadia M. Brashier, Robert Bond, Lia C. Bozarth, Ceren Budak, Ullrich K. H. Ecker, Lisa Fazio, Emilio Ferrara, Andrew J. Flanagin, Ales-sandro Flammini, Deen Freelon, Nir Grinberg, Ralph Hertwig, Kathleen Hall Jamieson, Kenneth Jo-seph, Jason Jones, R. Kelly Garrett, Daniel Kreiss, Shannon McGregor, Jasmine McNealy, Drew Margolin, Alice Marwick, FiIippo Menczer, Miriam Metzger, Seungahn Nah, Stephan Lewan-dowsky, Philipp Lorenz-Spreen,...
The identification of popular and important topics discussed in social networks is crucial for a better understanding societal concerns. It also useful users to stay on top trends without having sift through vast amounts shared information. Trend detection methods introduced so far have not used the network topology has thus been able distinguish viral from that are diffused mostly news media. To address this gap, we propose two novel structural trend definitions call coordinated...
This article investigates the prevalence of high and low quality URLs shared on Twitter when users discuss COVID-19. We distinguish between health sources, traditional news misinformation sources. find that misinformation, in terms tweets containing from websites, is at a higher rate than information websites. However, both are relatively small proportion overall conversation. In contrast, sources much rate. These findings lead us to analyze network created by referenced webpages users. When...
The spread of content produced by fake news publishers was one the most discussed characteristics 2016 U.S. Presidential Election. Yet, little is known about prevalence and focus such content, how its changed over time, this related to important election dynamics. In paper, we address these questions using tweets that mention two presidential candidates sampled at daily level, mentioned in tweets, open-ended responses from nationally representative telephone interviews. results our analysis...
Low-dimensional vector-space representations of academic periodicals reveal insights into their complex relationships.
The First Law of Geography states "Everything is related to everything else, but near things are more than distant things". This spatial significance has implications in various applications, trend detection being one them. In this paper we propose a new algorithmic tool, GeoScope , detect geo-trends. data streams solution that detects correlations between topics and locations sliding window, addition analyzing independently. offers theoretical guarantees for detecting all trending...
The rising prevalence of fake news and its alarming downstream impact have motivated both the industry academia to build a substantial number classification models, each with unique architecture. Yet, research community currently lacks comprehensive model evaluation framework that can provide multifaceted comparisons between these models beyond simple metrics such as accuracy or f1 scores. In our work, we examine representative subset classifiers using very set performance error analysis...
Past work has explored various ways for online platforms to leverage crowd wisdom misinformation detection and moderation. Yet, often relegate governance their communities, limited research been done from the perspective of these communities moderators. How is currently moderated in that are heavily self-governed? What role does play this process, how can process be improved? In study, we answer questions through semi-structured interviews with Reddit We focus on a case study COVID-19...
The success of a group depends on continued participation its members through time. We study the factors that affect user in context educational Twitter chats. To predict whether attended her first session particular chat will return to group, we build 5F Model captures five different factors: individual initiative, characteristics, perceived receptivity, linguistic affinity and geographical proximity. Through statistical data analysis thirty chats over two year period as well survey study,...
Do social movements actively shape the opinions and attitudes of participants by bringing together diverse groups that subsequently influence one another?Ethnographic studies 2013 Gezi uprising seem to answer "yes, " pointing solidarity among were traditionally indifferent, or even hostile, another.We argue two mechanisms with differing implications may generate this observed outcome: "influence" (change in attitude caused interacting other participants); "selection" (individuals who...
The spread of fake news on social media platforms has garnered much public attention and apprehension. Consequently, both the tech industry academia alike are investing increased effort to understand, detect, curb news. Yet, researchers differ in what they consider be sites. In this paper, we first aggregate 5 lists 3 mainstream sites published by experts reputable organizations. Then, focusing tweets about democratic (Hillary Clinton) republican (Donald Trump) nominees 2016 U.S....
Two recent disruptions to the online advertising market are widespread use of ad-blocking software and proposed restrictions on third-party tracking, trends that driven largely by consumer concerns over privacy. Both primarily impact display (as opposed search native social ads), affect how retailers reach customers content producers earn revenue. It is, however, unclear what consequences these are. We investigate using anonymized web browsing histories 14 million individuals, focusing...
Research on online political communication has primarily focused content in explicitly spaces. In this work, we set out to determine the amount of talk missed using approach. Focusing Reddit, estimate that nearly half all takes place subreddits host less than 25% time. other words, cumulatively, non-political spaces is abundant. We further examine nature and show conversations are toxic subreddits. Indeed, average toxicity comments replying a out-partisan even co-partisan replies
The framing of political issues can influence policy and public opinion. Even though the plays a key role in creating spreading frames, little is known about how ordinary people on social media frame issues. By new dataset immigration-related tweets labeled for multiple typologies from communication theory, we develop supervised models to detect frames. We demonstrate users' ideology region impact choices, message's influences audience responses. find that more commonly-used issue-generic...
Exposure to large language model output is rapidly increasing. How will seeing AI-generated ideas affect human ideas? We conducted an experiment (800+ participants, 40+ countries) where participants viewed creative that were from ChatGPT or prior experimental and then brainstormed their own idea. varied the number of examples (none, low, high exposure) if labeled as 'AI' (disclosure). Our dynamic design -- in condition are used stimuli for future same mimics interdependent process cultural...
Metaphor, discussing one concept in terms of another, is abundant politics and can shape how people understand important issues. We develop a computational approach to measure metaphorical language, focusing on immigration discourse social media. Grounded qualitative science research, we identify seven concepts evoked (e.g. "water" or "vermin"). propose evaluate novel technique that leverages both word-level document-level signals metaphor with respect these concepts. then study the...