Costas Mavromatis

Ph.D. Candidate at the Computer Science Department of University of Minnesota.

prof_pic.jpg

464 Walter Library

117 Pleasant St SE

Minneapolis, MN 55455

I am a 4th year Ph.D. Student at University of Minnesota, advised by Prof. George Karypis. I work on exciting projects with applications in question answering, text classification, and graph learning.

Before that, I obtained my 5-year Degree from University of Thessaly, Greece, advised by Prof. Manolis Vavalis.

I had two wonderful summers (2021, 2022) as an Applied Scientist intern at Amazon Web Services (AWS), CA.

My research interests span across natural language processing and graph machine learning. Some thechnologies that I am experinced with include Large Language Models, Graph Neural Networks, and Knowledge Graphs. Combining the above, my goal is to desing AI systems that are reliable, truthful, and accurate.

Useful links can be found at the bottom of this page.

news

Mar 15, 2022 Passed my Oral/Written Preliminary Exams with the work “Graph InfoClust: Leveraging cluster-level information for unsupervised graph representation learning”. Thanks George for the supervision, and the committee chairs (Prof. Georgios B. Giannakis, Prof. Zhi-li Zhang, and Prof. Gopalan Nadathur) for the feedback!
Aug 26, 2019 Accepted at the Ph.D. Computer Science program at University of Minnesota! Received the ADC (1-year) fellowship!
Jul 1, 2018 Successfully defended my MEng/BSc Thesis “Blockchain Design and Implementation for Decentralized Optimal Power Flow”! Thanks Prof. Manolis Vavalis and Dr. Magda Foti for the supervision!

selected publications

  1. EMNLP
    ReaRev: Adaptive Reasoning for Question Answering over Knowledge Graphs
    Costas Mavromatis, and George Karypis
    Findings of the Association for Computational Linguistics: EMNLP 2022, 2022
    PDF
  2. AAAI
    Tempoqr: temporal question reasoning over knowledge graphs
    Costas Mavromatis, Prasanna Lakkur Subramanyam, Vassilis N Ioannidis, and 5 more authors
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2022
    PDF
  3. PAKDD
    Graph infoclust: Maximizing coarse-grain mutual information in graphs
    Costas Mavromatis, and George Karypis
    In Advances in Knowledge Discovery and Data Mining: 25th Pacific-Asia Conference, PAKDD 2021, Virtual Event, May 11–14, 2021, Proceedings, Part I, 2021
    PDF