近來,由於網際網路的普及,許多專業領域知識亦逐漸資訊化,面臨這知識經濟的時代,人們唯有尋求最佳資訊科技工具,解決「資訊過量」的問題,進而改善資料檢索的效能。 由於生物醫學的資料採礦的技術日漸受到重視,故本研究將致力於利用資訊檢索配合搜尋引擎代理人技術,在龐大的生物醫學文獻資料庫中擷取出適當的資訊。 本研究擬針對「關鍵頁」、「疾病與基因資料分析」兩個方向進行從文獻中搜尋有用資訊的工作,其中關鍵頁檢索部分配合SimNet架構中之Matching-Degree計算模組進行文章相似度比對,並將搜尋伺服器建構在分散式處理的架構上,以增進檢索速度。本研究與Web畫面充分整合,除展現個人化資訊服務與資料檢索功能整合外,同時亦確保系統相容性。
In recent years, more and more biomedical-related information is available electronically due to the wide spreading of Internet and Web services. Mining valuable biomedical information from the literature has become an important issue. In the era of knowledge economy, in order to make the most value out of the information efficiently, it is important to apply the emerging information technologies to improve information overloading problem while searching information from the Internet. In this study, two search servers are proposed and implemented. The first one is a “Keypage Search Server” which finds similar documents from the Internet given a “keypage” with desired information. The second one is a “Genes-by-Disease Search Server” which finds related genes for a disease interested. The “matching-degree” module in SimNet architecture is applied in the keypage search server for measuring document similarity. Distributed computing mechanism is implemented in both servers to improve the computing performance as well as the network communication efficiency. Web-based architecture is used in this study to provide a user-friendly interface and ensure system compatibility.