This paper argues that norms, not individual choices, govern how young people use social media and proposes designing an independent platform to build trusted connections.

This paper conducts a systematic review of non-social media, free-text datasets for mental health research, revealing their predominant focus on English and depression detection, and identifying key g…

This paper proposes six indicators to improve bias detection in online news articles and tests their effectiveness in a two-phased experiment.

This paper introduces BOUTEF, a large-scale multilingual corpus for fake news in North Africa, and finds that fake news leverages emotionally charged, hybrid linguistic practices to enhance virality,…

TIBlender is a multi-agent system that integrates fragmented cyber threat signals from multiple social media platforms to generate comprehensive, actionable threat intelligence reports, significantly…

The paper demonstrates that generative AI can automate and scale highly personalized, context-aware spear-phishing attacks using only public social media data, resulting in messages that are significa…

The paper introduces SuiChat-CN, a novel Chinese group-chat benchmark for contextual suicide risk assessment, demonstrating that multi-party conversational context is crucial for accurate detection.

The paper introduces FigSIM, the first fine-grained dataset for analyzing suicide memes, which is used to benchmark models across tasks like suicide severity and figurative language detection.

This paper proposes HCIG and GCCN, two novel frameworks for multimodal sarcasm and cyberbullying detection using hierarchical cross-modal incongruity modeling and graph-based reasoning.

This paper compares sparse and dense retrieval models for scientific claim-source retrieval on the CheckThat! 2026 benchmark. Translating claims into English and incorporating publication metadata imp…

This paper conducted a randomized controlled trial on Reddit to test the effectiveness of various deescalation strategies in reducing personal insults using automated replies.

This study analyzes the dynamics of AI-generated multimodal misinformation using a large-scale dataset, finding that while synthetic content is highly viral, its spread is passive and its detectabilit…

The paper introduces TikStance, a multimodal and context-aware dataset for stance detection in political discussions on TikTok.

The authors propose a new solution for the content cold-start problem in industry-scale search and recommender systems, reducing bias, improving model prediction, and validating long-term impact.

This study explores why and how users customize social AI, identifying motivations and contributions to a closer relationship.

This paper systematically measured web tracking across 20 popular AI chatbots, finding that a majority share both conversational content and user identity information with third parties.

The paper empirically investigates the lead marketing ecosystem, revealing a highly non-compliant system that aggressively collects, shares, and monetizes sensitive personal data through deceptive bro…

This paper examines how ChatGPT is framed in educational YouTube videos and investigates the relationship between framings and audience response, using multimodal metadata and Epistemic Network Analys…

The paper proposes a hierarchical framework, PHF (Practice-Habitus-Field), inspired by Bourdieu's Theory of Practice, to improve LLM personalization by modeling user behaviors at three distinct levels…

The paper evaluates LLM-generated reactions to Spanish online news, finding that off-the-shelf models fail to accurately reproduce the measurable properties of real audience discourse, and even fine-t…