傳統的搜尋引擎根據使用者所提供的字詞進行網頁搜尋,但一般的搜尋引擎網頁排名機制,如:網頁排名(PageRank)、使用者點閱的多寡、廣告…等,很少個別使用者之使用習性,所以不同使用者在使用相同字詞時都會產生相同搜尋結果,只能根據關鍵字的相似度來進行推薦。
本研究將使用者的使用行為納入考量,動態重組搜尋引擎的搜尋結果,可以將使用者所需的結果優先列出;方便使用者挑選。在本研究動態機制中,使用者在使用搜尋引擎輸入關鍵字後,搜尋引擎提供搜尋結果回傳給使用者,讓使用者可以依照所搜尋的相關網站分類或是過去使用者的經驗中提供新的排序搜尋網站提供點選。本研究改進搜尋引擎,提供個人化的的動態搜尋結果給使用者進行網頁鏈結的瀏覽,以個人化動態的推薦點選率為評估依據,更容易的讓使用者搜尋到想要的相關網頁資訊。
The traditional search engines work by the key words supplied by users, but the ranking logic of results of general search engines, such as the page rank, clicks through rate, advertisement… etc, are infrequently based on users’ searching habits. Therefore, the search engines recommend pages only according to the similarity of key words because the different users will get the same results when they use the same key words..
The recommended searching system is built on the searching result dynamically. This novel system is processing a new order of result from the outcome, produced by regular searching engine, by either different category or the record user has searched.In this research, I was improving the regular searching engine for providing customized searching result dynamically. Based on the rate of link being chosen from the result produced from the system I designed, the users can be more effective to find the information they are looking for.