Payam Karisani
Update 2025: I have a new appointment at the department of Information Sciences. I will be helping with high-dimensional low-resource Machine Learning and Natural Language Processing tasks for automatically analyzing scientific documents to be used in a chatbot (partially sponsored by the US National Science Foundation).
I am Payam Karisani (it is a Kurdish name, I am Kurdish). I am a researcher in NLP and ML at the computer science department in the University of Illinois at Urbana-Champaign (UIUC). I work on large language models and their deficiencies. I have also done research on training models for low-data regime, I have published papers on Domain Adaptation, Active Learning, and Semi-Supervised Learning.
Selected academic activities (the list will not be updated anymore): pc-member for ICML 2020–2024, ICLR 2021–2024, NIPS 2021–2024, AAAI 2021–2024.
10. PropaInsight: Toward Deeper Understanding of Propaganda in Terms of Techniques, Appeals, and Intent (Sponsored by the US Department of Defense, Project SemaFor)
Venue: COLING 2025
Authors: J. L., L. A., Z. L., P. K., Z. H., M. F., P. N., J. H., H. J.
Link
9. Fact Checking Beyond Training Set (Sponsored by the US Department of Defense, Project SemaFor)
Venue: NAACL 2024 (full paper at the main research track, acceptance rate: 23%)
Authors: P. K., H. J.
Link | Code (soon)
8. Named Entity Recognition Under Domain Shift via Metric Learning for Life Sciences (Sponsored by the US Department of Defense, the US Department of Energy, and the US National Science Foundation)
Venue: NAACL 2024 (full paper at the main research track, acceptance rate: 23%)
Authors: H. L., Q. W., P. K., H. J.
Link
7. Masked LoGoNet: Fast and Accurate 3D Image Analysis for Medical Domain
Venue: KDD 2024 (full paper at the main research track, acceptance rate: 20%)
Authors: A. M., P. K., M. Z., S. C., N. D., H. J., S. P., R. R.
Link
6. Neural Networks Against (and For) Self-Training: Classification with Small Labeled and Large Unlabeled Sets
Venue: ACL 2023 (full paper at the Findigs research track, acceptance rate: ~30% out of 3872 papers.)
Authors: P. K.
Link | Code
5. Multi-View Active Learning for Short Text Classification in User-Generated Data (Sponsored by the US National Science Foundation)
Venue: EMNLP 2022 (full paper at the Findigs research track, acceptance rate: ~30% out of 3242 papers.)
Authors: P. K., N. K., L. X.
Link | Data
4. Multiple-Source Domain Adaptation via Coordinated Domain Encoders and Paired Classifiers
Venue: AAAI 2022 (full paper at the main research track, acceptance rate: 15.0% out of 9020 papers)
Authors: P. K.
Link | Code
3. Semi-Supervised Text Classification via Self-Pretraining
Venue: WSDM 2021 (full paper at the main research track, acceptance rate: 18.6% out of 603 papers)
Authors: P. K., N. K.
Link | Code
2. Domain-Guided Task Decomposition with Self-Training for Detecting Personal Events in Social Media (Sponsored by the US National Institutes of Health, and the US National Science Foundation)
Venue: WWW 2020 (full paper at the main research track, acceptance rate: 19.2% out of 1129 papers)
Authors: P. K., J. H., E. A.
Link
1. Did You Really Just Have a Heart Attack? Towards Robust Detection of Personal Health Mentions in Social Media
Venue: WWW 2018 (full paper at the main research track, acceptance rate: 15.0% out of 1140 papers. Best Paper Runner-Up.)
Authors: P. K., E. A.
Link | Code | Data