資訊檢索 (Information Retrieval)系統在人們的生活中已經是一個不可或缺的重要工具,相關回饋 (Relevance Feedback)的領域中Rocchio演算法因實作簡單且具有一定的效能,所以經常被廣泛的使用與研究,其概念為分析相關回饋結果中字詞的重要程度,作為挑選查詢擴展 (Query Expansion)字詞之依據,使資訊檢索系統更貼近使用者的資訊需求,但是其僅單純著重於字詞的出現頻率作為文件相關與否的依據,沒有考慮到字詞之間是否存在其他可以善加利用的資訊,而且在真實世界中出現頻率最高的字詞不一定與使用者的資訊需求具有相關性。因此本研究運用相關回饋的概念,分析相關文件中字詞之間所隱含的語意關係與共現關係,萃取其中適合的相關字詞作為查詢擴展之字詞來源,目的在於使查詢關鍵字能更符合使用者之資訊需求,解決過往相關回饋僅考慮字詞出現頻率而忽略的語詞資訊,並透過實驗證明本研究所提出之方法與其他方法相比之下,皆能有不錯的檢索效果,達到提升文件檢索準確率之最終目的。;Information retrieval systems are an indispensable tool in people′s lives. Rocchio’s query expansion method is simple and effective in the analyzing of the importance of terms residing in relevance feedback. However, in the make up of terms and its importance for query expansion, Rocchio’s method only focuses on term frequency and ignores other relationships between terms. Therefore, this study is aimed to develop a method in the utilization of the information of relevance feedback to analyze the semantic and co-occurrence relationships of terms in relevant documents to extract adequate relevant terms for query expansion. The results of experiments show that the proposed method of this study is effectiveness in document retrieval.