隨著網路資訊量的大量增加，檢索、擷取、學習和組織網路資訊和知識已經成為智慧型網際網路資訊系統所要達成的主要目標。DIKI搜尋法是整合資料探勘、智慧型代理人系統、資訊擷取、知識表達和資料庫系統等研究的跨研究領域。我們在論文計畫中，根據DIKI之相關研究領域，首先探討文字內容語意相關的重要研究問題；並根據DIKI方法，完成DIKI底層子系統雛形之模擬。期望後續能研究出自動擷取網站的有用深層資訊，以改善一般資訊和知識擷取之效能。 Along with a great deal of increment of the network information quantity, how to index ,retrieval , organize , and learn this network information and knowledge has become the most important research area of the intelligent internet information system .The DIKI search method is to study the integrate of different research areas. That are the data mining, the intelligent agent system, the information retrieval, the knowledge expression, and database system, etc. in the DIKI method research realm, we are in this thesis to find out the Related important research problems of inquiring into the character content language idea; we have also finished the simulation of the DIKI first floor subsystem prototype. The future research will focus on the automatic knowledge Retrieval of the useful deep information from the website to improve the retrieval capability of general information and knowledge.
