NADI 2026 TASK 5.2 SLU – SLOT FILLING

The Slot Filling subtask evaluates systems on their ability to extract semantic information directly from spoken utterances in Tunisian Arabic. Given an input audio recording, participants must produce the corresponding transcription with inline slot annotations following the SLURP-TN format. Submissions are evaluated using Concept Error Rate (CoER) and Concept-Value Error Rate (CVER). For more details, visit the **official NADI 2026 page**: https://nadi.dlnlp.ai/2026/ The **dataset** is available here: https://huggingface.co/datasets/Elyadata/SLURP-TN and **baselines** are available here: https://github.com/elyadata/SLURP-TN-baselines