一個針對網路服務語意檔案的自動標記方法

Abstract

網路服務的語意檔案通常會被描述成語意網路本體語言，透過此方法可以使用較一般化且機器可處理的語言描述網路服務的語意和功能。此外它還有助於解決語意網路的媒合、選擇和組合問題。然而標記網路服務語意檔案是一件困難且複雜的任務，因此本研究提出一個基於服務知識發現來做網路服務語意檔案的自動標記。首先將服務分別由使用者行為紀錄和網路服務檔案中取得服務被使用的紀錄以及輸入輸出等相關資訊；再將取出資訊使用關聯規則及輸入輸出媒合方法計算出結果來萃取出語意知識並產生相對應的網路服務語意檔案標記。本研究設計模擬實驗驗證可用性，經過不同的模擬實驗可證實透過關聯規則加上輸入輸出的媒合方法可找出網路服務之間的語意關聯。隨著使用者行為紀錄增加，平均準確度可提升至80%，平均召回率可提升至60%；而且隨著網路服務增加，平均召回率可提升到75%。

The Semantic Web approaches to Web services is often described in OWL-S. It gives the ability to describe the semantics of Web services and their capabilities in a formal and machine-processable manner. Moreover, it is helpful to semantic services matching, selection, and composition. However, annotating semantic Web services is a very heavy and complicated task. In this paper, this thesis proposed a methodology for knowledge discovery and automatic semantic annotation in Web services. With the proposed approach, semantic relations between Web services could be extracted by combining association rules and input/output matching. Then some simulations and results would be discussed to prove the availability of this methodology and also to show how this method could improve the recall for service matching. The results of simulations show that service matching precision and recall would be average improve to 80% and 60% when the number of user profiles increase. On the other hand, when the number of web services increase, this approach could improve average recall about 75%.