ARCADE: A City-Scale Corpus for Fine-Grained Arabic Dialect Tagging

2026-01-05

arXiv.org

AI Summary

This study addresses the lack of fine-grained, city-level annotations in existing Arabic multi-dialect datasets, which has hindered high-precision dialect identification research. To bridge this gap, the authors introduce ARCADE, the first city-scale Arabic speech corpus, comprising 3,790 broadcast audio segments from 58 cities across 19 countries and 6,907 multidimensional metadata annotations provided by native speakers, covering dialect, register, and emotion. By integrating automated streaming audio collection, 30-second segment extraction, and multi-native speaker validation, ARCADE achieves unprecedented fine-grained dialect labeling at the city level, encompassing both Modern Standard Arabic (MSA) and diverse dialectal varieties. The dataset is publicly available on Hugging Face, offering a high-quality benchmark for fine-grained dialect recognition and multi-task learning in Arabic speech processing.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

The Arabic language is characterized by a rich tapestry of regional dialects that differ substantially in phonetics and lexicon, reflecting the geographic and cultural diversity of its speakers. Despite the availability of many multi-dialect datasets, mapping speech to fine-grained dialect sources, such as cities, remains underexplored. We present ARCADE (Arabic Radio Corpus for Audio Dialect Evaluation), the first Arabic speech dataset designed explicitly with city-level dialect granularity. The corpus comprises Arabic radio speech collected from streaming services across the Arab world. Our data pipeline captures 30-second segments from verified radio streams, encompassing both Modern Standard Arabic (MSA) and diverse dialectal speech. To ensure reliability, each clip was annotated by one to three native Arabic reviewers who assigned rich metadata, including emotion, speech type, dialect category, and a validity flag for dialect identification tasks. The resulting corpus comprises 6,907 annotations and 3,790 unique audio segments spanning 58 cities across 19 countries. These fine-grained annotations enable robust multi-task learning, serving as a benchmark for city-level dialect tagging. We detail the data collection methodology, assess audio quality, and provide a comprehensive analysis of label distributions. The dataset is available on: https://huggingface.co/datasets/riotu-lab/ARCADE-full

Problem

• Arabic dialect
• city-level tagging
• speech corpus
• dialect identification
• fine-grained annotation

Innovation

• fine-grained dialect tagging
• city-level Arabic dialect
• speech corpus
• multi-task learning
• Arabic radio speech