Dhawaan, Waxaan Arkay 2 Warqadood oo Wanaagsan oo LLM+KG ah oo ku Saabsan Sababaynta Macquulka ah ee Adag

Dhawaan, Waxaan Arkay 2 Warqadood oo Wanaagsan oo LLM+KG ah oo ku Saabsan Sababaynta Macquulka ah ee Adag\n\n1. LARK https://arxiv.org/abs/2305.01157\nComplex Logical Reasoning over Knowledge Graphs using Large Language Models\n\n2. ROG https://arxiv.org/abs/2512.19092\nA Large Language Model Based Method for Complex Logical Reasoning over Knowledge Graphs\n\n## I. Dhibaatada Sababaynta Garaafyada Aqoonta\n\nGaraafyada Aqoonta (KG), oo ah sidaha udub dhexaadka u ah aqoonta qaabaysan, waxay wajahayaan saddex dhibcood oo xanuun leh:\n\n- Kakanaanta: Sababaynta badan, isgoysyada iyo ururada, diidmada iyo isku darka hawlgallada kale ayaa qarxaya\n- Dhameystir la'aan: KG-yada dhabta ah ee adduunka ayaa guud ahaan leh buuq iyo maqnaansho\n- Guud ahaan: Hababka dhexgalka dhaqameedka way adag tahay in laga guuro xog-ururin\n\nXaladaha dhaqameed (sida Query2Box, BetaE) waxay ku tiirsan yihiin boosaska dhexgalka joomatari, iyagoo u qaabeeyay hawlgallada macquulka ah sida vector/sanduuq xisaabeed, laakiin waxay lumiyaan macluumaad badan marka sababaynta qoto dheer. Sidee loo heli karaa moodel fahmaya qaab dhismeedka macquulka ah, sidoo kale si dabacsanaan leh u sababeeyaa? Kor u kaca moodellada luqadda weyn (LLM) ayaa bixiyay fikrado cusub.\n\n<img src=\