Matúš Pikuliak

Research areas: natural language processing, deep learning, transfer learning, language modeling, multilinguality

Position: Research Consultant 10/2022-01/2024

Email
Google Scholar
Research Gate
ORCiD
ResearcherID
LinkedIn
GitHub

Matúš focuses on natural language processing and deep learning, with a special interest in multilinguality, transfer learning and language modeling. He also works on issues related to online antisocial behavior, such as hate speech detection or misinformation detection, aiming to make these detection methods interpretable, trustworthy, and easy to train and deploy.

During his time at the university, he supervised more than ten bachelor’s and master’s theses, taught deep learning as a lead lab assistant, and organized focused academic seminars. He is a member of the ACL and Slovak.AI communities, and has participated in various research projects dealing with natural language processing, disinformation detection and recommender systems.

He was honored with the Award for exceptional diploma thesis from the Institute of Informatics of the Slovak Academy of Sciences’ and has experience consulting in industry projects utilizing deep learning neural models. In 2021, he obtained a Doctoral degree in Informatics from the Technical University of Košice.

Other notable projects

All publications

Selected Student Supervising

Master

  • Kňaze Adam – Transferability of Hate-Speech Classification Models Based on Deep Neural Networks. Defended 2020.
  • Vejčíková Veronika – Automatic detection of deceptive text style. Defended 2021.
  • Bojkovský Michal – Deep semi-supervised learning for misinformation detection in medicine articles. Defended 2021.

Bachelor

  • Vejčíková Veronika – Text generation with neural networks. Defended 2019.
  • Kňaze Adam – Text generation with neural networks. Defended 2018.
  • Bojkovský Michal – Multilingual neural networks. Defended 2019.