摘要 Abstract 

本次演講介紹結合知識的情境資訊檢索 (KCIR) 框架，旨在提升查詢上下文和檢索結果的品質。除了現有的方法，如情境路徑檢索、情境路徑生成和非單調情境路徑生成之外，我們還將討論 KCIR 未來的發展方向，包括運用不同種類的知識圖譜，來應對情境更豐富的檢索任務。

This talk presents a comprehensive framework for knowledge-enriched contextual information retrieval (KCIR), with a focus on integrating knowledge graphs (KGs) to enhance query context and retrieval outcomes. In addition to exploring current approaches like Contextual Path Retrieval (CPR), Contextual Path Generation (CPG), and Non-Monotonic Contextual Path Generation (NMCPG), we also discuss potential directions for KCIR development. These include leveraging temporal, commonsense, and causal knowledge graphs for more dynamic and contextually rich retrieval tasks.