NRK at FoRC 2024 Subtask I: Exploiting BERT-Based Models for Multi-class Classification of Scholarly Papers
摘要
This study presents the system developed by team NRK for SubTask I of the Field of Research Classification (FoRC) - NLSP 2024. The task focuses on single-label, multi-class classification of general scholarly papers. Our approach exploits the capabilities of various pre-trained BERTopology models, combined with a straightforward ensemble voting scheme to enhance classification performance. The proposed system achieved competitive results, ranking within the Top 5 on the final scoreboard.