Frank at NADI 2023 Shared Task: Trio-Based Ensemble Approach for Arabic Dialect Identification

Authors

Dilshod Azizov, Mohamed Bin Zayed University of Artificial IntelligenceFollow
Jiyong Li, Sun Yat-Sen University
Shangsong Liang, Mohamed Bin Zayed University of Artificial IntelligenceFollow

Document Type

Conference Proceeding

Publication Title

ArabicNLP 2023 - 1st Arabic Natural Language Processing Conference, Proceedings

Abstract

We present our system designed for Subtask 1 in the shared task NADI on Arabic Dialect Identification, which is part of ARABICNLP 2023. In our approach, we utilized models such as: MARBERT, MARBERTv2 (A) and MARBERTv2 (B). Subsequently, we created a majority-voting ensemble of these models. We used MARBERTv2 with different hyperparameters, which significantly improved the overall performance of the ensemble model. In terms of performance, our system achieved a competitive an F1 score of 84.76. Overall, our system secured the 5th position out of 16 participating teams.

First Page

637

Last Page

641

Publication Date

1-1-2023

Recommended Citation

D. Azizov et al., "Frank at NADI 2023 Shared Task: Trio-Based Ensemble Approach for Arabic Dialect Identification," ArabicNLP 2023 - 1st Arabic Natural Language Processing Conference, Proceedings, pp. 637 - 641, Jan 2023.