一個適用於行動裝置的網頁搜尋結果分群系統之研究

A Web Search Results On-the-fly Clustering System for Mobile Devices

平震宇(Chen-Yu Ping)

指導教授 ： 陸承志

元智大學/資訊學院/資訊管理學系/碩士(2007年)

https://doi.org/10.6838/YZU.2007.00211

摘要

雖然行動裝置呈現逐漸普及的趨勢，但對利用行動裝置存取網際網路的使用者來說，仍欠缺一個好用的行動搜尋方案，使得以行動裝置進行資訊搜尋不如使用個人電腦般便利。本研究依照行動裝置有限的運算能力和螢幕尺寸等特性，實做了一個網頁搜尋結果分群系統，這個系統採用一個遞迴式單階分群流程，每一次分群僅把特定網頁資料集做單階分群，並於分群完成後，使用語意傾向找出群集中具有語意之多個詞組，並從中選取最具代表性的詞組做為該群集之名稱，最後，再於行動裝置上以視覺化的方式呈現分群結果。透過視覺化的使用者介面，使用者可以快速的了解分群情形，並以點選的方式，瀏覽分群後之網頁搜尋結果。
　　本研究以空間密度對分群結果進行評估，實驗數據顯示，群集中之文章彼此相似，而群集間彼此並不相似。此外，分群結果及系統之使用者滿意度調查顯示，對於分群結果、系統之呈現方式、操作方式及整體表現，使用者均感到滿意。

關鍵字

行動裝置 ； 行動搜尋 ； 網頁搜尋 ； 文件分群 ； 語意傾向 ； 使用者介面

並列摘要

Using mobile device to access the Internet is getting increasingly popular, but there is still a lack of a satisfactory mobile search solution. Conducting web search on mobile device is currently not as easy as that on desktop PCs.
　　This study proposes and develops a web search results on-the-fly clustering system specially adapted to mobile devices features including limited computing power and screen size. The proposed system uses a recursive clustering method to group given web search results into one level clusters at a time. Once a clustering run is finished, several lexically related terms are chosen from each cluster as candidates of cluster label. Among them, the most representative term set (including up to three terms) is then chosen as the cluster label. Finally, the clustering results are shown on mobile devices in a graphical Context+Focus view. The GUI of this system allows user to view and manipulate clustering results easily.
　　We used space density to evaluate the clustering results. The experimental results shows that the intra-cluster density is high, while the inter-cluster density is low. Moreover, the results of two user studies reveals that users are satisfied with the cluster results and the system’s usability and performance.

並列關鍵字

Mobile Device ； Mobile Search ； Document Clustering ； Web Search ； Lexical Affinity ； Mobile User Interface

參考文獻

[3] Ball, G. and Hall, D.A. “A Clustering Technique for Summarizing Multivariate Data,” Behavioral Science, 12, pp. 153-155, 1967.

[7] Chittaro, L. “Visualizing Information in Mobile Devices,” Computer, March 2006, pp. 40-45.

[9] Fung, B.C.M., Wang, K, and Ester, M. “Hierarchical Document Clustering Using Frequent Itemsets,” In Proceedings of the SIAM International Conference on Data Mining. pp. 59-70, 2003.

[10] Jones, S., Jones, M., Deo, S. “Using Keyphrases as Search Result Surrogates on Small Screen Devices,” Pers Ubiquit Comput, 8, 2004, pp.55-68.

[11] Kummamuru, K., Lotlikar, R., and Roy, S. “A Hierarchical Monothetic Document Clustering Algorithm for Summarization and Browsing Search Results,” In Proceedings of International WWW Conference, New York, USA. 23 May 2004, pp.658-665.