Alexandria: A Multi-Domain Dialectal Arabic Machine Translation Dataset for Culturally Inclusive and Linguistically Diverse LLMs
This addresses the limited utility of translation tools for millions of Arabic dialect speakers, though it is incremental as it builds on existing dataset efforts.
The authors tackled the problem of machine translation systems performing poorly on Arabic dialects by introducing Alexandria, a large-scale dataset covering 13 countries and 11 domains with 107K samples, which serves as both a training resource and benchmark, exposing significant challenges in current models.
Arabic is a highly diglossic language where most daily communication occurs in regional dialects rather than Modern Standard Arabic. Despite this, machine translation (MT) systems often generalize poorly to dialectal input, limiting their utility for millions of speakers. We introduce \textbf{Alexandria}, a large-scale, community-driven, human-translated dataset designed to bridge this gap. Alexandria covers 13 Arab countries and 11 high-impact domains, including health, education, and agriculture. Unlike previous resources, Alexandria provides unprecedented granularity by associating contributions with city-of-origin metadata, capturing authentic local varieties beyond coarse regional labels. The dataset consists of multi-turn conversational scenarios annotated with speaker-addressee gender configurations, enabling the study of gender-conditioned variation in dialectal use. Comprising 107K total samples, Alexandria serves as both a training resource and a rigorous benchmark for evaluating MT and Large Language Models (LLMs). Our automatic and human evaluation of Arabic-aware LLMs benchmarks current capabilities in translating across diverse Arabic dialects and sub-dialects, while exposing significant persistent challenges.