{"slug": "cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations", "title": "CAMMAR: Culture-Aware Matryoshka for Metaphorical Arabic Representations", "summary": "Researchers introduce CAMMAR (Culture-Aware Matryoshka for Metaphorical Arabic Representations), a representation learning framework that organizes Arabic word meanings into nested lexical, cultural, and metaphorical embedding subspaces. Evaluated on a new span-annotated Arabic metaphor set, the geometric readout detects metaphor with AUC up to 0.84, with figurative meanings scoring higher than literal counterparts in 82.6% of pairs. The team will release datasets, cultural concept inventory, and code upon acceptance.", "body_md": "arXiv:2607.15847v1 Announce Type: new\nAbstract: Metaphor in Arabic is a culturally grounded mechanism for constructing meaning, encoding cultural knowledge that shapes interpretation. Yet current Arabic language models typically collapse lexical, cultural, and metaphorical information into a single representational space, a phenomenon we term \"semantic smearing\". We introduce CAMMAR (Culture-Aware Matryoshka for Metaphorical Arabic Representations), a representation learning framework that organizes meaning into nested lexical, cultural, and metaphorical embedding subspaces through a staged semantic curriculum. The design implements compositional principles of Al-Jurjani's theory of nazum, modeling figurative meaning as compositionally grounded in prior semantic relations, and yields a training-free geometric measure of metaphoricity based on the distance between lexical and metaphorical representations.\nEvaluated on a new span-annotated Arabic metaphor set as word-matched figurative/literal pairs, the geometric readout detects metaphor well above chance when the inter-layer geometry is shaped by paired supervision (AUC up to 0.84; figurative outscores its literal counterpart for the same word in 82.6\\% of pairs), but sits at chance under an unsupervised domain contrast alone, a clean separation between a legible-under-supervision regime and a non-emergent one. A controlled ablation shows that grounding the lexical layer in morphological roots gives a small but consistent gain, an effect absent from direct probing that reflects the layer's quality as a measurement anchor. We will release the datasets, cultural concept inventory, and code upon acceptance.", "url": "https://wpnews.pro/news/cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations", "canonical_source": "https://arxiv.org/abs/2607.15847", "published_at": "2026-07-20 04:00:00+00:00", "updated_at": "2026-07-20 13:41:36.777947+00:00", "lang": "en", "topics": ["natural-language-processing", "artificial-intelligence", "machine-learning"], "entities": ["CAMMAR", "Al-Jurjani"], "alternates": {"html": "https://wpnews.pro/news/cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations", "markdown": "https://wpnews.pro/news/cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations.md", "text": "https://wpnews.pro/news/cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations.txt", "jsonld": "https://wpnews.pro/news/cammar-culture-aware-matryoshka-for-metaphorical-arabic-representations.jsonld"}}