學術產出-國科會研究計畫

文章檢視/開啟

書目匯出

Google ScholarTM

政大圖書館

引文資訊

TAIR相關學術產出

題名 合作式個人化推薦系統之進階技術研究及其應用 (I)
其他題名 The Research of Advanced Techniques and Their Applications for a Collaborative Personalized Recommendation System
作者 陳良弼
貢獻者 國立政治大學資訊科學系
行政院國家科學委員會
關鍵詞 合作式個人化推薦系統
日期 2009
上傳時間 12-十一月-2012 11:03:49 (UTC+8)
摘要 伴隨著資訊科技之發展,各式各樣的物件在網際網路上迅速的累積,資訊過量成為主要的問題。因此,如何在大量的資料中,精確地推薦有用的資訊給使用者,即成為極具挑戰的研究課題。合作式推薦是其中一個解決資訊過量的方法;然而,隨著應用規模的成長,現階段合作式推薦系統所面臨的資料型態、處理模式與處理規模,都與過去單純的資料環境有著極大的不同,也導致現有技術有其侷限性。為了克服該問題,本計畫以三年為期研發一下世代合作式推薦系統。 在本年度計畫執行過程中,我們已完成預定完成之研究項目,分別為具不確定資料環境之子空間天際線查詢處理、高維度空間之任意子空間KNN查詢及具高準確率之鏈結樣式社群探索技術,並已發表於國際一流會議。本期中報告茲就本年度所完成的研究成果進行報告。
With rapid growth of the Internet technology, the information overloading starts to be a challenge. Therefore, an efficient and effec-tive approach to assist users to precisely get the useful information from the massive data-set is needed. The collaborative recommenda-tion mechanism is a popular solution to solve this problem. However, with the growth of the scale of applications, nowadays collaborative recommendation systems have to deal with dynamic and fast growing environments, in which the existing techniques become ineffi-cient and ineffective for high-quality recom-mendation results. Therefore, the advanced techniques for collaborative recommendation become important research issues to further study. In this progress report, we report three research results we achieved in this year. They are (1) efficient computation of sub-space top-k probabilistic skylines on uncertain data, (2) enhancing the accuracy of multimedia data retrieval by an online classifier, and (3) dis-covering link-pattern based communities by classical clustering methods.
關聯 應用研究
學術補助
研究期間:9808~ 9907
研究經費:1949仟元
資料類型 report
dc.contributor 國立政治大學資訊科學系en_US
dc.contributor 行政院國家科學委員會en_US
dc.creator (作者) 陳良弼zh_TW
dc.date (日期) 2009en_US
dc.date.accessioned 12-十一月-2012 11:03:49 (UTC+8)-
dc.date.available 12-十一月-2012 11:03:49 (UTC+8)-
dc.date.issued (上傳時間) 12-十一月-2012 11:03:49 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/55455-
dc.description.abstract (摘要) 伴隨著資訊科技之發展,各式各樣的物件在網際網路上迅速的累積,資訊過量成為主要的問題。因此,如何在大量的資料中,精確地推薦有用的資訊給使用者,即成為極具挑戰的研究課題。合作式推薦是其中一個解決資訊過量的方法;然而,隨著應用規模的成長,現階段合作式推薦系統所面臨的資料型態、處理模式與處理規模,都與過去單純的資料環境有著極大的不同,也導致現有技術有其侷限性。為了克服該問題,本計畫以三年為期研發一下世代合作式推薦系統。 在本年度計畫執行過程中,我們已完成預定完成之研究項目,分別為具不確定資料環境之子空間天際線查詢處理、高維度空間之任意子空間KNN查詢及具高準確率之鏈結樣式社群探索技術,並已發表於國際一流會議。本期中報告茲就本年度所完成的研究成果進行報告。-
dc.description.abstract (摘要) With rapid growth of the Internet technology, the information overloading starts to be a challenge. Therefore, an efficient and effec-tive approach to assist users to precisely get the useful information from the massive data-set is needed. The collaborative recommenda-tion mechanism is a popular solution to solve this problem. However, with the growth of the scale of applications, nowadays collaborative recommendation systems have to deal with dynamic and fast growing environments, in which the existing techniques become ineffi-cient and ineffective for high-quality recom-mendation results. Therefore, the advanced techniques for collaborative recommendation become important research issues to further study. In this progress report, we report three research results we achieved in this year. They are (1) efficient computation of sub-space top-k probabilistic skylines on uncertain data, (2) enhancing the accuracy of multimedia data retrieval by an online classifier, and (3) dis-covering link-pattern based communities by classical clustering methods.-
dc.language.iso en_US-
dc.relation (關聯) 應用研究en_US
dc.relation (關聯) 學術補助en_US
dc.relation (關聯) 研究期間:9808~ 9907en_US
dc.relation (關聯) 研究經費:1949仟元en_US
dc.subject (關鍵詞) 合作式個人化推薦系統en_US
dc.title (題名) 合作式個人化推薦系統之進階技術研究及其應用 (I)zh_TW
dc.title.alternative (其他題名) The Research of Advanced Techniques and Their Applications for a Collaborative Personalized Recommendation Systemen_US
dc.type (資料類型) reporten