教育部挑戰二○○八∼e世代人才培育計畫中，將「建構全民網路學習系統」列為其三大重點領域之一，目前已在各級學校建置完成計畫所需之硬體設備，期望藉由這些設備幫助使用者在網路上共創、共享教學資源，以促進網路學習之應用與發展。 雖然已有完善之設備，但在網路之利用上仍有未及之處，就現況而言，一般人在網路上尋找資料或尋求協助的方式大致有以下二種，一為至搜尋引擎輸入關鍵字，雖然回應速度很快，但往往查得大批無關資料，尚需一一檢視，花費許多時間才能得到想要之資料。二則於特定地方如網路論壇等尋求協助，雖然可避免上述問題而找到較合適的解答，但此種方式所需之等待時間較長。針對上述問題，本研究提出一種新的以本體論結合意圖為基之搜尋模式，使用者以自然語言輸入問題描述之後，搜尋引擎即可自動分析自然語言之語意，進而尋得符合使用者意圖之相關資料。 本研究搜尋資料之模式分為五個步驟進行，分別為建構數學領域本體、建構5W1H意圖句型庫、問句型態分析、意圖轉換、關鍵字擷取與擴展，最後將關鍵字組置於現有之搜尋引擎進行搜尋。 實驗結果顯示，本研究所提出之結合數學領域本體與5W1H意圖進行搜尋之方式，確實可提高搜尋資料之準確率並縮短查詢時間。 根據政府所做的一項調查，數學是許多學生懼怕，甚至討厭的學科，因此本研究以國小數學領域為主，希望在數學的教學與學習上均能幫助使用者獲得協助。 “Establish an Environment that supports Life-long Learning” is one of the three focus areas in the “Challenge 2008: E-Generation Manpower Cultivation Plan” proposed by the Ministry of Education. The project has implemented the required computer hardware in schools of all levels to facilitate the sharing and creation of online teaching resources, and develop and promote the application of online learning. Currently, the way most users search information online can be categorized into two types: 1. Entering keywords into search engines. Normally users receive responses from search engines in a short time, but much of the retrieved information is irrelevant. Thus they need to spend much time in reviewing the search results item by item to obtain the information they really need. 2. Seeking information or helps from certain places online, such as internet forums. Information obtained this way is more likely to be what the users wanted but it takes longer to receive responses. It seems that each of these two search methods has its advantages and disadvantages. Therefore, we propose a new way to search information on the internet, namely the natural language search method. Users seeking information can enter their problem description in natural language format into a search engine. The search engine automatically analyzes the nature of the problem to use this information to speed up the search process. Thus it will be able to quickly search the database and return appropriate information to users. The five main steps of this research are: developing ontology for mathematics, structuring intention-sentence pattern database, transforming intention, identifying and expanding keywords, and finally including keyword sets in the search engine to improve the searching process. Computational results show that combining mathematics ontology with 5W1H-intention can improve the accuracy of the search and shorten the search time. Therefore, it can help users obtain the information they desired on the internet efficiently and effectively. According to a government’s survey, mathematics is the subject that most students fear or even dislike. For this reason, our research focuses on the applications of the proposed methods to the mathematics curriculum of elementary schools. We hope this method can help users in the teaching and learning of mathematics
