基於使用者存取模式的網站結構最佳化

隨著網路應用的蓬勃發展，網站的結構也越來越複雜，使用者往往必須花費釵h的瀏覽時間才能找到所需的資料。所謂的適應性網站，就是透過網頁探勘技術發掘使用者的瀏覽需求，以改進網站的結構和呈現方式，增進使用者的瀏覽效率。方法之一是根據個別使用者的需求，提供額外的輔助導覽機制供其參考。另一種方法則是重整網站的結構以適應多數使用者的瀏覽需求。但目前的網站重整方法並未考慮重整後的結構是否為整體最佳的結構。因此，本研究提出一個網站重整的0-1線性規劃模式，根據探勘使用者的瀏覽行為模式所得到的網頁連結機率，建立最佳化的網站結構。此模式除能滿足大多數使用者的瀏覽需求外，並可限制網頁對外的超連結數目及與首頁的距離，避免使用者的資訊過載及過多的點選。使用0-1線性規劃模式雖然可以得到最佳解，但是當網站結構過於複雜時將耗費甚多求解時間。於是本研究另外發展一個螞蟻演算法以降低求解網站結構最佳化問題的時間。實驗結果顯示，本研究的螞蟻演算法具有相當高的求解效率與效能。實例探討的結果顯示，本研究所提之方法的確能在有限超連結及搜尋深度的限制下，獲得滿足大多數使用者瀏覽需求的網站結構。

關鍵字

螞蟻演算法；網站結構重整；網頁使用探勘；使用者瀏覽型態；適應性網站； 0-1線性規劃

並列摘要

As the Internet and thus Web sites grow at an astounding rate, Web site structures become more and more complicated. Web site visitors often have to spend significant time surfing the Web site to retrieve their required information. Based on Web usage mining, adaptive Web sites dig user browsing patterns from Web server logs so as to automatically improve their structures and presentations. One of two directions for improving Web sites is to facilitate a specific user by providing additional guidance; the other involves modifying the Web site to ease the navigation for a large set of users. Nevertheless, the latter modifies Web site locally rather than globally such that the new structure is not necessarily a global optimum. This study proposes 0-1 programming models for reorganizing Web sites based on the cohesion between Web pages obtained through Web usage mining. Besides, the proposed models reduce the user information overload and search depth. The 0-1 programming models are able to obtain the optimal solutions; however, the required computation time increases rapidly with the Web site scale. Therefore, this study also proposed an ant colony system to reduce the computation time. The results with numerical examples suggest that the ant colony system significantly reduce the computation time while giving high-quality solutions. An empirical study also shows that the proposed ant colony system is capable of providing Web site structures that can meet the requirements of most users under limited hyperlinks and search depth.

並列關鍵字

0-1 linear programming ； Web site reorganization ； Adaptive Web site ； Web usage mining ； User browsing pattern ； Ant colony system

參考文獻

[1] M. Perkowitz and O. Etzioni, “Adaptive web sites: an AI challenge”, Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence, Nagoya, Japan, 1997, pp. 16-23.

[2] R. Cooley, B. Mobasher and J. Srivastava, “Data preparation for mining world wide web browsing patterns,” Knowledge and Information Systems, Vol. 1, 1999, pp. 5-32.

[4] S. Hollfelder, V. Oria and T. Ozsu, “Mining user behavior for resource prediction in interactive electronic malls,” Proceedings of 2000 IEEE International Conference on Multimedia and Expo, NY, USA, Vol. 2, 2000, pp. 862-866.

[10] M. Perkowitz and O. Etzioni, “Towards adaptive web sites: conceptual framework and case study,” Artificial Intelligence, Vol. 118, 2000, pp. 245-275.

[12] B. Mobasher, R. Cooley and J. Srivastava, “Creating adaptive web sites through usage-based clustering of URLs,” Proceedings of the 1999 IEEE Knowledge and Data Engineering Exchange Workshop, Chicago, Illinois, 1999.

國際替代計量

基於使用者存取模式的網站結構最佳化

全文下載

主題瀏覽