While social media make it easy to connect with and access information from anyone, they also facilitate basic influence unfriending mechanisms that may lead segregated polarized clusters known as "echo chambers." Here we study the conditions in which such echo chambers emerge by introducing a simple model of sharing online networks two ingredients unfriending. Users can change both their opinions connections based on are exposed through sharing. The dynamics show even minimal amounts...

Deepfakes are synthetic content generated using advanced deep learning and AI technologies. The advancement of technology has created opportunities for anyone to create share deepfakes much easier. This may lead societal concerns based on how communities engage with it. However, there is limited research available understand perceive deepfakes. We examined deepfake conversations Reddit from 2018 2021 -- including major topics their temporal changes as well implications these conversations....

Online social media are increasingly facilitating our interactions, thereby making available a massive "digital fossil" of human behavior. Discovering and quantifying distinct patterns using these data is important for studying behavior, although the rapid time-variant nature large volumes make this task difficult challenging. In study, we focused on emergence "collective attention" Twitter, popular networking service. We propose simple method detecting measuring collective attention evoked...

Bird songs are acoustic communication signals primarily used in male-male aggression and male-female attraction. These often monotonous patterns composed of a few phrases, yet some birds have extremely complex with large phrase repertoire, organized non-random fashion discernible patterns. Since structure is typically associated function, the structures bird provide important clues to evolution animal systems. Here we propose an efficient network-based approach explore structural design...

The Moral Foundations Dictionary (MFD) is a useful tool for applying the conceptual framework developed in Theory and quantifying moral meanings implicated linguistic information people convey. However, applicability of MFD limited because it available only English. Translated versions are therefore needed to study morality across various cultures, including non-Western cultures. contribution this paper two-fold. We first Japanese version (referred as J-MFD) using semi-automated method—this...

Moral foundations theory explains variations in moral behavior using innate foundations: Care, Fairness, Ingroup, Authority, and Purity, along with experimental supports. However, little is known about the roles of relationships between those everyday situations. To address these, we quantify from a large amount online conversations (tweets) topics on social media site Twitter. We measure loadings latent semantic analysis tweets related to abortion, homosexuality, immigration, religion,...

Abstract An infodemic is an emerging phenomenon caused by overabundance of information online. This proliferation makes it difficult for the public to distinguish trustworthy news and credible from untrustworthy sites non-credible sources. The perils debuted with outbreak COVID-19 pandemic bots (i.e., automated accounts controlled a set algorithms) that are suspected spreading infodemic. Although previous research has revealed played central role in misinformation during major political...

Abstract The anti-vaccine movement has gained traction in many countries since the COVID-19 pandemic began. However, their aggressive behaviour through replies on Twitter—a form of directed messaging that can be sent beyond follow-follower relationships—is less understood, and even is known about language use differences this behaviour. We conducted a comparative study anti-vaxxers’ behaviours by analysing longitudinal dataset tweets English Japanese. found two common features across these...

Although warning interventions are increasingly used to curb the spread of misinformation, their effectiveness varies significantly depending on implementation. This study online survey experiments examine how timing and frequency warnings affect sharing intention dubious videos. Participants evaluated videos featuring politicians they support made decisions about social media. In Study 1 (n = 558), a label was presented either during or after watching single video; in 2 631), participants...

The rise of generative AI has raised societal concerns about the potential synthetic media or deepfakes for misinformation. While previous research established that negative emotions drive text-based misinformation sharing, their influence on deepfake dissemination remains poorly understood. Through two online experiments (N = 487; N 479), we examined how emotional valence in images affects sharing intentions across political and entertainment news domains. Using GPT-3 Stable Diffusion,...

The rise of generative AI has raised societal concerns about the potential synthetic media or deepfakes for misinformation. While previous research established that negative emotions drive text-based misinformation sharing, their influence on deepfake dissemination remains poorly understood. Through two online experiments (N = 487; N 479), we examined how emotional valence in images affects sharing intentions across political and entertainment news domains. Using GPT-3 Stable Diffusion,...

Unlike simple biological rhythms, the rhythm of oscine bird song is a learned time series diverse sounds that change dynamically during vocal ontogeny. How to quantify development one most important challenges in behavioural biology. Here, we propose method, called ‘rhythm landscape’, visualize and how structure, which measured as durational patterns silences, emerges changes over development. Applying this method Bengalese finch songs, show structure begins with broadband develops into...

The COVID-19 pandemic brought about several challenges in addition to the virus itself. rise of Islamophobic hate speech on social media is one such challenge. As countries were coping with economic collapse due mass lockdown, hateful people, especially those associated far-right groups, targeting and blaming Muslims for spread coronavirus. In India, where intense religious/communal polarization taking place under right-wing Bharatiya Janata Party (BJP)-led government, prominent instance...

The popularity of the instant messaging app Telegram in Ukraine and Russia was already high before still-ongoing Russian invasion Ukraine. However, since 24 February 2022 (when began), it has seen huge increases subscribers even become primary communication news source In this exploratory research, we analyzed channels from both (@UkraineNow — official channel Ukrainian government, @V_Zelenskiy_official Volodymyr Zelenskyy) (@rt_russian network RT) to understand patterns Ukraine-Russia...

The Linguistic Inquiry and Word Count Dictionary 2015 (LIWC2015) is a standard text analysis dictionary that quantifies the linguistic psychometric properties of English words. A Japanese version LIWC2015 (J-LIWC2015) has been expected in fields natural language processing cross-cultural research. This study aims to create J-LIWC2015 through systematic investigations original corpora. entire was initially subjected human machine translation into Japanese. After verifying frequency use words...

QAnon is an umbrella conspiracy theory that encompasses a wide spectrum of people. The COVID-19 pandemic has helped raise the to wide-spreading movement, especially in US. Here, we study users' dynamics on Twitter related movement (i.e., pro-/anti-QAnon and less-leaning users) context infodemic topics involved using simple network-based approach. We found pro- anti-leaning users show different population late were mostly anti-QAnon. These trends might have been affected by Twitter's...

False information spreads on social media, and fact-checking is a potential countermeasure. However, there severe shortage of fact-checkers; an efficient way to scale desperately needed, especially in pandemics like COVID-19. In this study, we focus spontaneous debunking by media users, which has been missed existing research despite its indicated usefulness for countering false information. Specifically, characterize the tweets with information, or fake tweets, that tend be debunked Twitter...

Using machine learning algorithms, including deep learning, we studied the prediction of personal attributes from text tweets, such as gender, occupation, and age groups. We applied word2vec to construct word vectors, which were then used vectorize tweet blocks. The resulting vectors inputs for training models, accuracy those models was examined a function dimension size results showed that algorithms could predict three interest with 60-70% accuracy.