基於個人本體之查詢邏輯建構及關鍵字推薦

Personal Ontology-based Query Logic Construction and Keyword Recommendation

蔡秉栓(Bing-shuan Tsai)

指導教授 ： 李思堯

亞洲大學/資訊學院/資訊多媒體應用學系碩士班/碩士(2011年)

摘要

網路上的資訊量迅速增加，使用者利用搜尋引擎檢索資訊時，若是無法明確指定關鍵字，將會造成檢索結果難以滿足其需求。過去的研究中，多是透過建立使用者輪廓(User Profile)的方式輔助使用者進行查詢擴展，但是這些研究只能解決使用者長期或短期資訊需求的其中一種。此外，大部分的查詢擴展系統只利用AND邏輯運算子組合查詢關鍵字，忽略了這些關鍵字彼此之間可能存在不同的語意關係。
本研究提出一套查詢建構輔助系統。此系統可根據使用者輸入之查詢關鍵字，利用個人本體推薦與長期資訊需求相關之關鍵字；同時，亦可根據使用者有興趣之檢索結果網頁，推薦與短期資訊需求相關之關鍵字。藉此，協助使用者隨時根據其當時之需求—無論長期或短期—取得所需之檢索結果。
本研究同時在使用者查詢的過程中，透過個人本體找出各查詢關鍵字間的語意關係，據此從AND、OR或NOT中選用合適的邏輯運算子組合查詢關鍵字，再轉換成搜尋引擎可以理解之邏輯敘述。最後，本研究亦驗證此演算法之可行性並實作成系統。

關鍵字

個人本體 ； 查詢邏輯建構 ； 關鍵字推薦 ； 查詢擴展

並列摘要

As the amount of information across the internet increases, users may not get the needed information if they could not specify the keywords precisely. Most of previous studies proposed query expansion to help users in their query process. Browsing records and some other data of a user was recorded in an individual user profile to reflect the information needs of the specific user. However, the user profile can usually represent either long-term or temporal interest of the user only. Moreover, most of the proposed query expansion method used AND operator to formulate the query without considering other possible semantic relationships among keywords.
This paper proposed a system to help users constructing their queries. Giving a query from the user, the system recommends related keywords based on the user’s personal ontology, which related to his long-term interest. The proposed system can also recommend keywords by analyzing the user designated web pages, which represent the temporal interest of the user. In this way, the users can retrieve the required web pages according to their information needs, related to either long-term or temporal interest, in a single system.
The proposed system also considered different semantic relationships among keywords based on personal ontology. Appropriate logic operators are selected to represent these relationships. Specifically, OR and NOT operators are also used in addition to AND operator. Then, an algorithm is applied to formulate and convert the query into a browser-acceptable format. Finally, the proposed system is implemented and the correctness and feasibility of the proposed algorithm are validated.

並列關鍵字

personal ontology ； query logic construction ； keyword recommendation ； query expansion

參考文獻

[23]蕭清標、楊豐兆，「具視覺化自動查詢擴展之網路影像檢索代理人系統」，第十一屆電子化企業經營管理理論暨實務研討會論文集，臺灣，2010。

[1]van Rijsbergen, C. J., Information Retrieval, retrieved June 03, 2011 from http://www.dcs.gla.ac.uk/Keith/Preface.html, 1979.

[7]Zhai, D., and Wang, L., "Study on Key Techniques of Query Expansion Based on Ontology and Its Application ", in Proceedings of the International Conference on Computational Intelligence and Software Engineering, China, 2009.

[10]Sahib, N. G., Tombros, A., and Ruthven, I., "Enabling Interactive Query Expansion through Eliciting the Potential Effect of Expansion Terms", in Proceedings of the 32nd European Conference on IR Research, UK, 2010.

[14]吳政穎，「利用相關回饋建立概念化的使用者興趣檔以協助使用者進行網頁查詢」，中央大學資訊管理學系，碩士論文，2009。