<ul><li>Large language models (LLMs) have made progress in reasoning over knowledge graphs.</li><li>Current methods suffer from a high non-retrieval rate, reducing accuracy in answering questions based on these graphs.</li><li>To overcome this issue, the concept of super-relations is introduced, enabling both forward and backward reasoning by summarizing and connecting relational paths within the graph.</li><li>The proposed ReKnoS framework improves retrieval efficiency and reasoning performance, demonstrating superior performance over existing baselines.</li></ul>

Reasoning of Large Language Models over Knowledge Graphs with Super-Relations

Discover more