[[alternative]]Design of an Integrated Mechanism of Search Engines on Web by Domain Ontology

[[abstract]] 近來由於產業及科技的競爭，以致於相關知識的蒐集、獲取、整合、儲存、管理、分享與運用之重要性相對提升。隨著網際網路發展，如何以自動化的方式有效獲取網路上的資訊提供使用者所需的知識是一項很大的挑戰。 本研究結合利用資料探勘發掘網頁內容知識並檢視其相似性且導入領域實體概念，發展強化搜尋引擎的過濾及排序機制，透過演算法去除格式不完整、有重覆性網址且針對格式化的摘要及標題進行資訊含量之運算，其值若介於本研究所設立之可接受範圍，便進一步計算摘要權重值；若遇到描述不同但意思相仿的摘要，會應用領域實體所建立的法則計算詞彙相似程度，其後給予適當權重值，本研究的領域實體是著重於國小數學學習方面，系統則將每篇摘要之權重排列順序，其後檢視符合原意與否，再取回其網頁內容，經由擷取就變成可利用知識，此知識可提供給使用者解決問題之參考。希冀能節省使用者自行過濾檢索時間與減少頻寬資訊量。[[abstract]] Due to the rapid development of information technology, it is important to search, gain, integrate, store, share, reuse and manage the different scopes of professional knowledge. This issue becomes increasingly essential for users to extract the appropriate internet information efficiently and automatically for a great variety of resources on websites. The research includes both web content mining and information retrieval to design a strengthened mechanism of search engines on web by domain ontology. We try to develop and design the algorithms which have the functions of filtering, ranking and weighting. The purpose is to filter the dump link and the advertisement link according to the web document titles, the ranking of the abstract’s URL and the weighting of the information content. Then the users can retrieve more suitable information and capture the web content knowledge efficiently. In the process of filtering and ranking, the knowledge can be refined to useful one which can offer end users to decide whether or not the knowledge meets their demands. In this way, the users can save the time to filter and retrieve and decrease the loading of internet