Nauka

Istraživanje

Istraživanja su usmerena na metode obrade prirodnih jezika, pronalaženja informacija i klasifikacije dokumenata zasnovane na mašinskom učenju i znanju, od n-gram modela i analize sentimenta do jezičkih resursa za srpski i primene velikih jezičkih modela u obrazovanju.

Oblasti istraživanja

Obrada prirodnih jezika
Veštačka inteligencija
Mašinsko učenje
Analiza sentimenta i emocija
Kategorizacija teksta
Jezički resursi za srpski
Primena velikih jezičkih modela u obrazovanju

Izabrani noviji radovi

2026
EduGrader: The AI-powered grading assistant
S. Iković, J. Graovac
Computer Science and Information Systems (ComSIS), OnLine-First, Issue 00, pp. 26-26
DOI
2026
Building an emotion lexicon for Serbian using curated language resources
M. Šošić, J. Graovac, R. Stanković
Language Resources and Evaluation, vol. 60, no. 1, art. 9
DOI
2026
A multilingual sentence-transformer baseline for Serbian newspaper topic classification
S. Petalinkar, M. Ikonić Nešić, R. Stanković, J. Graovac
International Conference on Language Technologies for Low-resource Languages (LATE-LRL), Fes, Morocco
2026
Same pace, different path: An NLP-based analysis of gender disparities in IT careers
M. Gligorijević, M. Vujošević Janičić, J. Graovac
Artificial Intelligence Conference, Belgrade, Serbia
2022
Meet your email sender - hybrid approach to email signature extraction
J. Graovac, I. Tomašević, G. Pavlović-Lažetić
Asian Conference on Intelligent Information and Database Systems (ACIIDS 2022), LNAI, Springer, pp. 547-558
DOI
2022
Effective methods for email classification: Is it a business or personal email?
M. Šošić, J. Graovac
Computer Science and Information Systems (ComSIS), vol. 19, no. 3, pp. 1155-1175
DOI

Kompletna bibliografija →

Uređivačke i druge aktivnosti

2018-
Pridruženi urednik (Associate Editor) časopisa Intelligent Data Analysis (IOS Press)
2022
Član programskog odbora, IPROSE 2022, ACIIDS 2022 (LNAI, Springer)
2022
Pozvano predavanje, NMLSP 2022: Numerical Methods for Large Scale Problems, Beograd