Shaden Shaar

Research Interests

• Multi-modal long-form generation . Long-form question-answering and summarization over videos and other multi-modal narratives, with an emphasis on coherence across extended outputs.
• Clinical NLP . Using LLMs to surface clinical decision patterns from unstructured medical narratives, with a focus on heart failure and heart transplant care.
• Automated fact-checking prior . Detecting, verifying, and justifying claim veracity across multi-modal, variable-length inputs.
• Propaganda detection prior . Identifying propaganda and persuasion techniques across multi-modal inputs in news articles, memes, and other media.
• COVID-19 misinformation prior . Analyzing misinformation, vaccine-related fake news, and the broader infodemic on social platforms during the COVID-19 pandemic.

Highlights

• h-index 24 with ~2,858 citations on Google Scholar.
• Best Demo Award (Honorable Mention) at ACL 2020 for Prta, a system for analyzing propaganda techniques in news.
• University Fellowship at Cornell University (2021) for exceptional preparation and promise.
• University Honors at Carnegie Mellon University (2019); 50% Academic Merit Scholarship at CMU (2015); five-time Dean’s List awardee.
• Co-organized multiple shared tasks at CLEF–CheckThat! (2020, 2021) and SemEval (2021) on automated fact-checking and propaganda detection.

Work Experience

• May 2025 — Aug 2025

  Applied Scientist Intern

  Zillow Group · Remote, USA

  • Built conversational assistive AI agents for Zillow's real-estate platform using reinforcement-learning methods.
  • Designed reward-modeling and evaluation pipelines for grounded multi-turn property-search dialogue.

• Jan 2025 — May 2025

  Machine Learning Research Engineer Intern

  ScaleAI, Inc. · New York City, NY, USA

  • Built an Arabic-language AI assistant for the Qatari judicial department to aid judges reviewing active cases.
  • Implemented dense and sparse retrieval over Arabic cassation and supreme-court rulings for relevant-precedent lookup.
  • Designed Arabic case-summarization and event-extraction pipelines feeding the judge-facing platform.

• May 2022 — Aug 2022

  AI/ML Research Intern

  Apple · Seattle, WA, USA

  • Worked with Dr. Alex Churchill on zero-shot multi-turn conversation data generation.
  • Built a model that generates conditioned multi-turn dialogue datasets for downstream training.

• Jul 2019 — Aug 2021

  Research Assistant

  Qatar Computing Research Institute (QCRI), HBKU · Doha, Qatar

  • Worked with Dr. Preslav Nakov and Prof. Giovanni Da San Martino on fact-checking and propaganda detection.
  • Introduced the task of detecting previously fact-checked claims; methods adopted by leading fact-checking organizations.
  • Built fact-checking and analysis pipelines for COVID-19 misinformation across social platforms.
  • Co-organized shared tasks at CLEF–CheckThat! (2020, 2021) and SemEval–2021 Task 6 on persuasion in memes.
  • Built and shipped Prta, a public propaganda-analysis demo (ACL 2020 Best Demo, Honorable Mention).

Selected recent and impactful publications. Full list on the publications page or Google Scholar.

• 
  Thematic Analysis of Accepted Exception Requests for Heart Transplant Candidates Using a Large Language Model

  J. Frye, Shaden Shaar, C. Cardie, E. DeFilippis, D. Estrin, G. Sayer, N. Uriel, et al.

  JHLT · 2026 · Journal

  Uses an LLM to perform thematic analysis of accepted exception requests for heart transplant candidates, surfacing the clinical rationales that drive decisions in a setting where manual review at scale is infeasible.

  Clinical
• 
  MovieRecapsQA: A Multimodal Open-Ended Video Question-Answering Benchmark

  Shaden Shaar, B. Thymes, S. Chaixanien, C. Cardie, B. Hariharan

  CVPR · 2026 · Conference

  An open-ended video-QA benchmark built from movie recaps that stress-tests whether models can reason over long-form narrative, not just short clips. Paired with baselines that expose a large gap between human and model performance on grounded, cross-modal questions.

  Long-form Generation
• 
  Are Triggers Needed for Document-Level Event Extraction?

  Shaden Shaar, W. Chen, M. Chatterjee, B. Wang, W. Zhao, C. Cardie

  TACL · 2025 · Journal · 2 citations

  Revisits a long-standing assumption in event extraction — that explicit trigger annotations are required — and shows that trigger-free formulations can match or exceed trigger-based pipelines at the document level.

  Events
• 
  Assisting the Human Fact-Checkers: Detecting All Previously Fact-Checked Claims in a Document

  Shaden Shaar, N. Georgiev, F. Alam, G. Da San Martino, A. Mohamed, P. Nakov

  EMNLP · 2022 · Findings · Conference · 45 citations

  Scales fact-checked-claim detection from isolated sentences to full documents, where each claim must be located and matched jointly. Introduces a document-level dataset and retrieval+ranking system tuned for real fact-checker workflows.

  Fact-checking
• 
  Prta: A System to Support the Analysis of Propaganda Techniques in the News

  G. Da San Martino, Shaden Shaar, Y. Zhang, S. Yu, A. Barrón-Cedeño, P. Nakov

  ACL · 2020 · Conference · 93 citations

  An end-to-end system for highlighting 18 propaganda techniques in news articles, paired with a public web interface. Recognized with an Honorable Mention for Best Demo at ACL 2020.

  Propaganda
• ACL 2020
  That Is a Known Lie: Detecting Previously Fact-Checked Claims

  Shaden Shaar, G. Da San Martino, N. Babulkov, P. Nakov

  ACL · 2020 · Conference · 241 citations

  Formalizes "previously fact-checked claim detection" as a ranking task and releases the first dataset for it, showing that reusing existing fact-checks is a practical alternative to verifying every claim from scratch.

  Fact-checking

Education

• Aug. 2021 — Present

  Ph.D. in Computer Science

  Cornell University · Ithaca, NY, USA

  Minor: Applied Mathematics.

• Aug. 2021 — May 2024

  M.S. in Computer Science

  Cornell University · Ithaca, NY, USA

• Aug. 2015 — May 2019

  B.S. in Computer Science

  Carnegie Mellon University · Doha, Qatar

  Minor in Mathematics; University Honors

Teaching Experience

Academic Services

• 2021
  Co-organizer. SemEval–2021 Task 6: Detection of Persuasion Techniques in Texts and Images.
• 2021
  Co-organizer. CLEF–2021 CheckThat! Lab on detecting check-worthy and previously fact-checked claims (Bucharest, Romania).
• 2020
  Co-organizer. CLEF–2020 CheckThat! Lab on automatic identification and verification of claims.
• 2020
  Reviewer. ACL, EMNLP, NAACL, EACL, COLING, AAAI, CIKM (Program Committee).

Contact

Happy to hear from potential collaborators or anyone curious about the work. The best way to reach me is by email: sshaar31@gmail.com. A printable CV is available as a PDF.