一個結合TOPSIS與AHP的網路服務信任評估機制

Abstract

隨著網路服務所提供的功能越來愈多且複雜，其資訊安全的問題也日益受到重視，目前網路服務的安全多集中於傳輸的安全性上，並未考慮到服務內容與安全性以外的問題。有鑑於此，本論文提出之“網路服務信任評估機制(Web Services Trust Evaluation Mechanism, WSTEM)”，選擇適當的網路服務評估準則(Criteria)，同時考慮主觀與客觀的安全因子，以使用者的安全要求為基礎，結合理想解類似度偏好順序評估調整法(Modified Technique for Order Preference by Similarity to Ideal Solution Method, TOPSIS)與層級分析法(Analytic Hierarchy Process, AHP)於網路服務信任值計算，不同於以往結合TOPSIS與AHP之應用，我們所提出的機制，考慮到使用者的主觀意見，加入使用者回饋與修正。在系統模擬實驗中，本論文模擬一可支援WSTEM之信任網路服務推薦系統(Trusted Web Services Recommendation System)，由模擬結果可發現，在惡意服務比例為10%、50%時，應用WSTEM與未應用WSTEM相比，正確推薦達100%，在惡意服務比例高達90%時，推薦錯誤率僅10%，故本機制可有效推薦具安全性的網路服務予使用者。

In recent years, the Web Services (WS) is becoming more complex and popular, and the security issues are also important. However, the WS Security standards only support the WS transaction security in the transmission, do not consider the WS content and other issue, such as reliability and so on. In this thesis, we propose a novel Web Services Trust Evaluation Mechanism (WSTEM) which uses appropriate WS evaluation criteria and considers the security factors in subjective and objective. WSTEM is based on the user’s WS-Security requirement, which uses Modified Technique for Order Preference by Similarity to Ideal Solution Method (TOPSIS) and Analytic Hierarchy Process (AHP) to compute the trust value of WS, it is different from other applications also use TOPSIS and AHP. Because it considers user preferences, adds user feedback and trust value revision for trusted WS recommendation. The simulation result shows that WSTEM outforms to the orginal study, by increasing the proportion of insecure service is 10% and 50%, that are recommend correct. When the proportion of insecure service is 90%, the recommend error rate is just 10%. Therefore, WSTEM can recommend secure and trusted services to users.