電機資訊學院: 資訊網路與多媒體研究所指導教授: 李允中王凱昇Wang, Kai-ShengKai-ShengWang2017-03-062018-07-052017-03-062018-07-052016http://ntur.lib.ntu.edu.tw//handle/246246/275730近年來本體論被認為是設計更精確和易於使用的資訊檢索、推薦系統以及預測引擎的有力策略，這已經成為一種突出的方式以提供一個符合用戶喜好信息，我們的系統藉由利用本體論以改善服務預測的精準度來做開發。 藉由透過個性化建立一個更準確的用戶輪廓，在預測服務可獲得一些額外的好處，當中包括更高的機會獲得用戶他們所期望的結果以及進一步為他們提供更好的整體體驗，因此我們決定將個人本體論結合了服務的描述擴展為服務本體論。 本論文旨在於給一個情境，我們想要預測它的相關Web服務，有四個主要核心模組方法被提出，分別為(1)分析情境到實體；(2)從實體轉為一個複雜事件；(3)建構服務本體論；以及(4)利用複雜事件和服務本體論來預測相關相關Web服務。 這些被提出方法的效果不但可以顧及到在服務預測過程中的情境架構，也可獲得服務預測的一些回饋，而這些回饋依賴於服務本體論，因此它可隨著時間持續學習提供更準確的預測。Nowadays, ontology is considered a powerful strategy for designing more precise and easy to use information search, recommendation system and prediction engine. It has become a prominent approach to providing information that is tailored to user preferences. Our system was developed by utilizing the ontology to improve the precision of service prediction. By building a more accurate profile of user through personalization, there are a number of benefits that can be gained in predicting service, including a higher chance of obtaining the results users expect, and an overall improved experience. Therefore, we decided to extend personal ontology with the description of services to a service ontology. The objective of this thesis is: give a scenario, and we would like to predict its related web services. There are four main modules in proposed approach: (1) parsing scenarios to entities; (2) transferring entities to a complex event; (3) constructing service ontology; and (4) using complex events and service ontology to predict its related web services. The benefits of proposed approach not only consider the structure of a scenario in the process of service prediction but also obtain the feedback of service prediction. The feedback depends on the service ontology, therefore it can provide more accurate predictions as time goes by.10924956 bytesapplication/pdf論文公開時間: 2021/8/3論文使用權限: 同意有償授權(權利金給回饋學校)本體論相似度頻率塑模建構演化情境複雜事件服務預測Web服務ontologysimilarityfrequencymodelingconstructionevolutionscenariocomplex eventservice predictionweb servicethesis10.6342/NTU201601595http://ntur.lib.ntu.edu.tw/bitstream/246246/275730/1/ntu-105-R03944040-1.pdf