ORCA: A Challenging Benchmark for Arabic Language Understanding

TL;DR

ORCA is a comprehensive benchmark designed to evaluate Arabic language understanding across diverse dialects and tasks, addressing the lack of such resources and enabling better measurement of progress in Arabic NLP.

Contribution

This work introduces ORCA, the first extensive Arabic language understanding benchmark covering multiple varieties and tasks, with a unified evaluation metric and public leaderboard.

Findings

ORCA covers 60 datasets across 7 NLU task clusters.

It enables comparison of 18 multilingual and Arabic models.

Provides a public leaderboard for Arabic NLU progress.

Abstract

Due to their crucial role in all NLP, several benchmarks have been proposed to evaluate pretrained language models. In spite of these efforts, no public benchmark of diverse nature currently exists for evaluation of Arabic. This makes it challenging to measure progress for both Arabic and multilingual language models. This challenge is compounded by the fact that any benchmark targeting Arabic needs to take into account the fact that Arabic is not a single language but rather a collection of languages and varieties. In this work, we introduce ORCA, a publicly available benchmark for Arabic language understanding evaluation. ORCA is carefully constructed to cover diverse Arabic varieties and a wide range of challenging Arabic understanding tasks exploiting 60 different datasets across seven NLU task clusters. To measure current progress in Arabic NLU, we use ORCA to offer a comprehensive comparison between 18 multilingual and Arabic language models. We also provide a public leaderboard with a unified single-number evaluation metric (ORCA score) to facilitate future research.