Towards Figurative Language Understanding in Moroccan Darija: An Evaluation of Large Language Models

Key points

• Figurative language remains a major challenge for Machine Translation (MT) systems, as the intended meaning of such expressions often goes beyond literal translation.1
• Although recent advances in Large Language Models (LLMs) improved the translation of Arabic dialects, benchmark resources for evaluating figurative language translation in Moroccan Darija remain scarce.1
• Using MDFC, we evaluate 19 open- and closed-weight language models and MT systems at both the expression and contextual levels under zero-shot and six-shot settings.1
• We also conduct layer-wise similarity-based probing to determine how model representations distinguish figurative from literal meanings.1

Sources (1)

• [1]Towards Figurative Language Understanding in Moroccan Darija: An Evaluation of Large Language Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 18, 05:05 PM

  To this end, we introduce the Moroccan Darija Figurative Language Corpus (MDFC), a benchmark dataset for MT and multiple-choice question (MCQ) selection covering idioms, metaphors, and proverbs.

  Figurative language remains a major challenge for Machine Translation (MT) systems, as the intended meaning of such expressions often goes beyond literal translation.

Extractive summary: sentences quoted from the sources.