網際網路服務推薦系統

A Web Services Recommendation System

吳緯閔(Wei-Min Wu)

指導教授 ： 蘇木春

國立中央大學/資訊電機學院/資訊工程學系/碩士(2008年)

摘要

由於網頁資訊藉著分散式架構的網路環境快速擴散，造成網路上的資訊越來越龐大。在這樣的環境下，使用者取得最適當的網際網路服務不會是一件容易的事，因此為了協助人們快速找尋適合的網際網路服務，資訊過濾技術的引用便快速地發展起來，而推薦系統便是資訊過濾裡頭的一種智慧型網路技術。所以本論文便將推薦系統應用在網際網路服務的推廣和資訊獲得上。
一般來說推薦系統主要分成三種：協同式推薦系統和內容式推薦系統，還有混合兩種推薦系統的優點的混合式推薦系統。本論文設計一個混合式推薦系統，是以協同式推薦系統為主，以內容式推薦系統來彌補其不足。
由於協同式推薦系統需要對物品作分群，現行網際網路服務分群的技術，仍需要系統維護者手動分群，這樣的方式不但效率不佳，而且掺入人為因素。所以本論文提出了一個利用服務敘述之全自動網際網路服務分群機制，其中包括了自動分群機制，自動分群命名機制，自動化新增網際網路服務等。經實驗証實，本論文所提出之全自動網際網路服務分群的正確率可以達到80%，爾後利用本論文提出全自動網際網路服務分群機制所產生之分群，應用在推薦系統上。使用者滿意度亦有7.3分的高滿意度。

關鍵字

網際網路服務 ； 分群命名 ； 服務分群 ； 推薦系統

並列摘要

In a large-scale distributed network environment like Internet, information has been increased and changed continuously. Accessing information in such dynamically changing, heterogeneous and world-wide distributed environments puts a big burden on the users. A possible solution to alleviate information overload is the use of recommendation systems. Recommendation Systems are a kind of web intelligence techniques to make daily information filtering for people.

In this paper, a web services recommendation system is proposed to help users to quickly retrieve the web services needed by them. To implement the web services recommendation system, we first develop a two-level clustering algorithm to automatically cluster web services into several groups. In addition, we propose a method to automatically search the most common characteristics from the services belonging to the same cluster to name the corresponding cluster. Based on the clustering results, appropriate Web services can then be effectively and quickly recommended to users. A possible solution to the “cold-start” problem is also implemented in the recommendation system. Simulation results demonstrated the performance of the proposed web services recommendation system is encouraging.

並列關鍵字

Services Clustering ； Web Services ； Keywords: Recommendation System

參考文獻

[2] M. Balabanovic and Y. Shoham, “Fab: content-based, collaborative recommendation,” Communications of the ACM, vol. 40, no. 3, pp. 66-72, March 1997.

[9] F. Curbera, M. Duftler, R. Khalaf, W. Nagy, N. Mukhi, and S. Weerawarana, “Unraveling the web services web: an introduction to SOAP, WSDL, and UDDI,” IEEE Internet Computing, vol. 6, pp. 86-93, April 2002.

[11] B. Esfandiari, M. Weiss, and Y. Luo, “Towards a classification of web service feature interactions,” Computer Networks, vol. 51, pp. 359-381, February 2007.

[12] C. F. Eick, N. Zeidat, and Z. Zhao, “"Supervised clustering" algorithms and benefits,” in Proceedings of the 16th IEEE International Conference on Tools with Artificial Intelligence, USA, November 2004, pp. 774-776.

[15] S. M. Li, C. H. Chi, C. Ding, S. Chen, and Y. Huang, “Automatic recommendation of quality requirements for software services,” in Proceedings of the 2005 Australian Software Engineering Conference, Australian, August 2005, p. 117.