News

Our work "Leadership as Coordination Control: Behavioral Signatures and the Recovery-Advantage Boundary in Multi-Agent LLM Teams" is accepted at TMLR (Transactions on Machine Learning Research).

Our work "XChoice: Explainable Evaluation of AI-Human Alignment in LLM-based Constrained Choice Decision Making" is accepted at AACL 2026 Findings.

Our work "SynSym: A Synthetic Data Generation Framework for Psychiatric Symptom Identification" is presented at KDD 2026 in Jeju.

Two of our papers accepted at EMNLP 2026 Findings: "Same Symptoms, Different Severity? Investigating Gendered Cue Effects on LLMs in Depression Severity Assessment"; and "Understanding Moral Reasoning Trajectories in Large Language Models: Toward Probing-Based Explainability".

Our work "Gendered cultural discourse in Japan across the prewar–postwar transition: evidence from historical word embeddings" is published in EPJ Data Science.

Our paper "When AI Becomes Routine: A Decade of Public AI Mediation in Korean Go Commentary" (extended version on arXiv) accepted at AIES 2026 (AAAI/ACM Conference on AI, Ethics, and Society).

Three of our papers accepted at ACL 2026: "PluRule: A Benchmark for Moderating Pluralistic Communities on Social Media" (main conference); "Vulnerability of LLMs' Belief Systems? LLMs Belief Resistance Check Through Strategic Persuasive Conversation Interventions" (Findings); and "Somatic in the East, Psychological in the West?: Investigating Clinically-Grounded Cross-Cultural Depression Symptom Expression in LLMs" (C3NLP workshop).

Three parallel talks and one poster from the lab were accepted at IC2S2 2026 (International Conference on Computational Social Science).

Two posters by our visiting students presented at CHI 2026: "When Behavior Is Not Enough: Identity Biases in Human-AI Creative Collaboration" and "If I Change Your Eyes, Are You Still You? Identity Thresholds in Deepfake Perception".

Congratulations to Rasika Muralidharan and Andrea Chung on passing their qualifying exams!

Sumit Kumar (IIIT-Hyderabad) and Jihyun Kim (Sungkyunkwan University) join the lab as visiting students this spring. Welcome!

Our work "A Cross-Cultural Comparison of LLM-based Public Opinion Simulation: Evaluating Chinese and U.S. Models on Diverse Societies" is presented at the IEEE ICDM 2025 workshops.

Welcome to Juyoung An and Shirui (Olivia) Zhou, who join the lab as PhD students this fall!

Hyunseon Won (Sungkyunkwan University) joins the lab as a visiting student this fall. Welcome!

Our work "A semantic embedding space based on large language models for modelling human beliefs" is published in Nature Human Behaviour! Read it here: https://rdcu.be/epoQJ

We are part of the research team studying "BRAIN (Belief Resonance and AI Narratives): Understanding Belief-Narrative Resonance in the Era of Generative AI." This five-year, $7.5 million project is funded by the U.S. Department of Defense.

Chaewon Kang and Migyeong Kang (Sungkyunkwan University) join the lab as visiting students this fall. Welcome!

Our work "X-posing Free Speech: Examining the Impact of Moderation Relaxation on Online Social Networks" is presented at the 8th Workshop on Online Abuse and Harms (WOAH) at NAACL 2024.

Our work "A Survey on Predicting the Factuality and the Bias of News Media" is accepted at ACL 2024 Findings.

Our work "What is Twitter, a Social Network or a News Media?" (WWW'10) has received 10,000 citations!

Our work "Rematch: Robust and Efficient Matching of Local Knowledge Graphs to Improve Structural and Semantic Similarity" is accepted at NAACL 2024 Findings.

Our work "The impact of toxic trolling comments on anti-vaccine YouTube videos" is published in Scientific Reports.

Our work "ChatGPT Rates Natural Language Explanation Quality Like Humans: But on Which Scales?" is accepted at LREC-COLING 2024.

Our work "Public Perception of Generative AI on Twitter: An Empirical Study Based on Occupation and Usage" has been published in EPJ Data Science.

Our work "Is ChatGPT better than Human Annotators? Potential and Limitations of ChatGPT in Explaining Implicit Hate Speech" (WWW Companion, 2023) has already got 100 citations!

Our work "Enhancing Spatio-temporal Traffic Prediction through Urban Human Activity Analysis" has been published in ACM CIKM'23.

Jisun serves as a workshop chair for the Web Conference 2024 and Haewoon serves as a PhD Symposium chair.

Our work "Can we trust the evaluation on ChatGPT?" is presented at TrustNLP 2023, co-located with ACL.

Our studies "Is ChatGPT better than Human Annotators? Potential and Limitations of ChatGPT in Explaining Implicit Hate Speech" and "Chain of Explanation: New Prompting Method to Generate Higher Quality Natural Language Explanation for Implicit Hate Speech" are accepted at The Web Conference 2023 poster track.

Our studies "Political Honeymoon Effect on Social Media: Characterizing Social Media Reaction to the Changes of Prime Minister in Japan" and "Wearing Masks Implies Refuting Trump?: Towards Target-specific User Stance Prediction across Events in COVID-19 and US Election 2020" got accepted for ACM WebSci'23.

After two amazing years at SMU, we moved to Luddy School of Informatics, Computing, and Engineering, Indiana University Bloomington.

Our work '"This is Fake News": Characterizing the Spontaneous Debunking from Twitter Users to COVID-19 False Information' is accepted at ICWSM'23.

Jisun serves as a PC chair of ASONAM 2022.

Haewoon serves on the best paper committee at WebSci'22.

Our work "Storm the Capitol: Linking offline political speech and online Twitter extra-representational participation on QAnon and the January 6 insurrection" is accepted in Frontiers in Sociology.

Our work "Modeling Political Activism around Gun Debate via Social Media" is accepted in ACM Transactions on Social Computing.

Our work "You Have Earned a Trophy: Characterize In-Game Achievements and Their Completions" is accepted at WebSci'22.

Our work "Measuring 9 Emotions of News Posts from 8 News Organizations across 4 Social Media Platforms for 8 Months" is published in ACM Transactions on Social Computing.

Our papers "Who Is Missing? Characterizing the Participation of Different Demographic Groups in a Korean Nationwide Daily Conversation Corpus" and "Understanding Toxicity Triggers on Reddit in the Context of Singapore", are accepted at ICWSM'22.

Our work "What really matters?: characterising and predicting user engagement of news postings using multiple platforms, sentiments and topics" is published in Behaviour & Information Technology.

Haewoon serves as a Virtual Experience chair of ACM FAccT (Fairness, Accountability, and Transparency) 2022.

Our work "Precision Public Health Campaign: Delivering Persuasive Messages to Relevant Segments Through Targeted Advertisements on Social Media" is published in JMIR Formative Research.

Jisun serves as a PC chair of AAAI ICWSM (International Conference on Web and Social Media) 2022.

Our work on anti-Asian slurs during COVID-19 is accepted at EMNLP 2021 Findings.

Our work "Estimating Homophily in Social Networks Using Dyadic Predictions" is published in Sociological Science.

Our work "FrameAxis: characterizing microframe bias and intensity with word embedding" is published in PeerJ Computer Science.

Haewoon was awarded the D.S. Lee Foundation Fellowship for a period of one year.

Jisun served as a PhD Symposium Chair at WebSci.

Jisun received the best reviewer award at ICWSM.

Our work "How-to Present News on Social Media: A Causal Analysis of Editing News Headlines for Boosting User Engagement" is presented at ICWSM.

Haewoon received the best reviewer award at The Web.

Our work "How to be successful in social media? A causal analysis of 8 media's news sharing practices on Twitter" is selected as an oral presentation in Computation+Journalism Symposium 2021.

Jisun serves as an associate editor of EPJ Data Science starting from 2021.