MetaSwarm at AbjadMed: Forensic Optimization and Class-Balanced Discovery for Medical Diglossia in Abjad Scripts

Read full text Summarise with VIRAT
Authors: Rahul Jaisy
Cited by: 0
References: 5

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The classification of diglossic medical text presents a high-dimensional challenge defined by extreme class imbalance (N = 82) and the orthographic ambiguity of unvocalized Abjad scripts.While standard supervised learning often collapses into majority-class prediction due to the "Long Tail" distribution, we introduce a Human-in-the-Loop Forensic Optimization framework.Unlike static end-to-end pipelines, our approach decouples strategic hyperparameter tuning from high-throughput tactical execution (Elastic Compute).We leverage a rigorous Class-Balanced Focal Loss (CBFL) derived from the "Effective Number of Samples" theory (E n ) to stabilize the decision manifold against stochastic class dominance.Using a CAMELBERT-DA backbone optimized via a custom weighted trainer on Dual H200 GPUs, our system achieved a robust Public Leaderboard score of 0.3588.We further perform a "Linguistic Error Topology" analysis, utilizing UMAP projections and attention saliency, to demonstrate that generalization gaps are driven by dialectal "Constraint Drift" rather than stochastic model failure.

Fields of Study & Subjects

Handwritten Text Recognition Techniques Imbalanced Data Classification Techniques Natural Language Processing Techniques

Publication details

Journal / source
Workshop on NLP for Languages Using Arabic Script (AbjadNLP)
Pages
144–148
Publication date
1 January 2026
Language
EN