Cross-topic stance detection in Arabic is the task of identifying whether a text expresses a pro, against, or neutral position toward a given issue, and it is particularly challenging under topic shifts and class imbalance. In Subtask B of the StanceNakba 2026 shared task on Arabic cross-topic stance detection, we are given a Levantine Arabic sentence and one of two topics: “Normalization with Israel” or “Refugee/Immigrant Presence in Jordan,” and we must classify the expressed stance. A central difficulty is the systematic failure of standard fine-tuning to recognize the minority neutral class, driven by majority-class dominance in cross-entropy training and accuracy-based checkpoint selection. To address this, we combine random oversampling with class-weighted cross-entropy loss, and we build an ensemble of four Arabic pre-trained transformers MARBERT, AraBERT Large, XLM-RoBERTa Base, and CAMeL-BERT Mix each trained using Stratified 5-Fold cross-validation. Our final system achieves a macro-F1 of 0.9777 and an accuracy of 97.79% on the evaluation set.

We describe the ZHR system for both tracks of StanceEval-2026, Arabic stance detection on Mawqif-XT, where Track 1 evaluates on seen targets and Track 2 on unseen ones. The system combines fine-tuned Arabic encoders, a few-shot classifier that retrieves labelled examples for a served instruction model, and generative models adapted with LoRA, and turns their blended probabilities into labels with a threshold rule that abstains to None unless one stance class is probable enough, with thresholds fixed on the development split. We report the whole sequence of experiments, including eight that lowered F_avg2 and were not adopted. The main finding is that transfer to a new target is carried by the demonstrations, the labelled examples retrieved into the prompt, rather than by the model itself. Generated on-target tweets in the example pool gave the largest gain on both tracks, matching the surface form of the test tweets helped where the target has a distinctive style, and doubling the examples hurt. Our official results are F_avg2 0.896 with accuracy 0.855 on Track 1, and F_avg2 0.940 with accuracy 0.922 on Track 2. The strongest official baselines reach 0.8425 on the Track 1 development set and 0.7475 on the Track 2 test set.

Two-phase Whisper fine-tuning (broad dialect pre-training + low-LR target annealing) reaches 57.0 WER on NADI 2026 Subtask 1.1, with 8 WER points from the schedule alone.

We describe our participation in Subtask 2 of the DialectSentEval 2026 Shared Task, whose goal is to rewrite an Arabic dialectal sentence so its sentiment polarity is reversed while its meaning, entities, and dialect are preserved. Our system, SWAPPER (Sentiment-SWAPping with Parameter-Efficient adapteRs), is an ensemble of LoRA-adapted mBART-50 checkpoints in which the adapters are extended beyond attention into the feed-forward sub-layers, where the actual lexical rewriting is learned. Three design choices carried most of the weight: (i) a minimal Arabic instruction prefix that reserves encoder capacity for content; (ii) feed-forward LoRA; (iii) forcing the decoder’s beginning-of-sequence token to ar_AR so mBART-50 does not drift into other scripts. On the official blind test SWAPPER reaches Sentiment-Style Accuracy 0.6732, BLEU 41.38, and chrF 65.84. As a secondary contribution, we introduce the J-Gap (𝛥 J = J_tgt - J_src), a Jaccard-based copy-collapse diagnostic that cleanly separates genuine sentiment swappers from models that merely copy their input, and explains why superficially strong backbones such as AraT5v2 fail the task despite high surface metrics.

Lexicons are indispensable tools for textual analysis through labelled term associations.Despite their utility, lexicons are static and require manual effort to curate and maintain.Regular updates are essential to stay relevant amid semantic shifts and neologisms during language evolution.In this work, we explore the potential of supervised learning for expanding lexicons using dictionaries.We study the effect of using dictionaries with varying properties such as noise, size, labels, structure and curation method.Definitions are used as input features to a transformer model (BERT) that assigns categories to terms.We analyse the expansions using varying English dictionaries and lexicons for estimated accuracy, coverage and labelling consistency and apply the expanded versions to a downstream task.Our analyses show dictionary based expansion is a robust approach.We release our expanded lexicons, code, and pretrained models.