透過您的圖書館登入
IP:3.144.113.197
  • 學位論文

利用情境脈絡協助關鍵字查詢的自動完成

Automatic Concept and Context Based Query Completion

指導教授 : 鄭卜壬

摘要


關鍵字的自動完成是各個搜尋領域的重要應用之一。它在檔案搜尋、網路搜尋,甚致是整合型程式碼編輯器,都有著重要的應用。它可以預測使用者的關鍵字並省下使用者的鍵盤輸入。 由於網路的急速成長以及移動裝置的成熟,在移動裝置上進行網絡搜尋是件很重要的事。亦因為手機等移動置的輸入介面很有限,沒有像個人電腦的硬體鍵盤,所以在網路搜尋上提供關鍵字的自動完成就變得更重要了。 在這本篇論文中,我們會建立一個以情境概念和脈絡為基礎的模型去改進關鍵字的自動完成。它會使用非監督式分群法去找出搜尋引擎記錄檔中關鍵字的不同概念,以及衡量關鍵字之間的聯繫性。然後根據使用者最近使用過的關鍵字,配合概念的變動而推薦最可能的詞組作為自動完成。這個模型還可以利用替代性法則,用來產生過往沒有在搜尋引擎中出現的詞組,以減輕搜尋在建立模型時的資料不足。在實驗部份,我們會以真正的搜尋記錄檔作為檢測,以此表現模型的可行性。我們還比較過去的自動完成的方法,結果顯示我們的方法有著明顯的改進。

並列摘要


Query completion is an application in many search domains, such as file search, programming editor word suggestions or the web search. It is used to save the user keystroke in every domain. As the rapid growth of the web and the increasing popular of mobile devices, searching web data from mobile device is an important issue. Therefore, query completion in web search is becoming more and more important because it can save users’ effort from the hard inputting mobile device. In this paper, we construct a concept and context base model to help to improve the query completion. It uses a clustering method to construct the concept of query and measure the dependences of queries in search log. This model also can generate the unseen query from log by the replaceable criteria to alleviate the sparseness problem. We experiment the model in real query log data to demonstrate the feasibility. We compare it with context base method proposed in the past. It also gets a significant improvement.

參考文獻


[1] Z. Bar-Yossef and N. Kraus. Context-Sensitive Query Auto-Completion. In WWW 2011.
[2] M. Barouni-Ebrahimi and A. Ghorbani. On Query Completion in Web Search Engines Based on Query Stream Mining. In WI 2007.
[3] H. Bast and I. Weber. Type Less, Find More: Fast Autocompletion Search with a Succinct Index. In SIGIR 2006.
[4] S. Bhatia, D. Majumdar and P. Mitra. A Novel Approach for Frequent Phrase Mining in Web Search Engine Query Streams. In SIGIR 2011.
[5] S. Bhatia, D. Majumdar and P. Mitra. Query Suggestions in the Absence of Query Logs. In SIGIR 2011.

延伸閱讀