近五次調查(94年7月至97年1月)結果得知，「搜尋資訊」為目前使用寬頻網路使用者最常使用之功能，而目前最常見的搜尋資訊方式，以性質來可區分為「主動」及「被動」兩種搜尋模式，主動性質的搜尋模式例如:直接於搜尋引擎中輸入關鍵字查詢，優點為反應即時，缺點為可能連同無關資料一併搜尋，因而需額外的資料過濾時間;被動性質性質的搜尋模式:例如至論壇留言等待回應，優點為所獲得的回應較為精準，缺點為回應時間非即時性，或可能缺乏該方面專家解答。 本研究針對兩種模式其優劣勢，提出一以5W1H意圖為基之醫學本體搜尋模式，自動解析並分類處理搜尋者搜尋時所使用的自然語言問句，找出所屬之意圖後對應所歸屬之意圖以及領域知識本體進行語意之擴展，將擴展關鍵字提供搜尋者於進一步於搜尋引擎中使用。 實驗結果顯示，本研究所提出透過5W1H意圖為基及醫學領域知識本體結合而進行搜尋的模式，能有效協助搜尋者得到所需醫學知識領域資料，並使搜尋資料之精確度提升及減少搜尋資料時間。 根據(USRDS)2007年報公佈最新全球尿毒症人口數排名，台灣洗腎病患佔全國總人口數比及每年新增之洗腎病患比，兩者皆為全球之首。網路上相關醫療資訊甚多，因此本研究針對慢性腎臟病領域建立其醫學領域知識本體，期望能對有相關醫療知識需求之搜尋者有所助益，進而提升醫療品質。 According to the five researches founded lately (July 2005 to January 2008), ＂searching information＂ is the most commonly function used by broadband network users. Ways to searching information can be divided into two ways, which are ＂active way＂ and ＂passive way＂. ＂Active way＂ means that the users input keywords into a search box for getting information. Advantage to this way is that users can get responses soon, while disadvantage is low precision of one's searches. ＂Passive way＂ means that people post questions to forums and waiting for feedbacks. Getting more precisely answers is a major advantage of this passive search, yet users may not capable to control the quality of the answers. The author proposed a search style based on 5W1H-intention with medical area ontology. This search engine can analysis and clarify searchers' natural language automatically; it can also define searchers' intentions and extend meanings of the phrases that searchers input. All the extended-keywords can also be provided to searchers for their search. Computational results of this research show that search styles which combine 5W1H-intention and medical area ontology might help information searchers to get medical information efficiently; moreover, it can improve precision of searches as well as reducing searching time. According to the annual report of USRDS in 2007, Taiwan's dialysis patients ratio to its population and new dialysis patients are the top of the world. Since too much medical information exists on the Internet, the author has tried to build medical area ontology of chronic kidney disease for helping information searchers as well as improving medical quality.
