旅遊語意網整體服務系統之建置

Abstract

如何有效的利用網際網路上的資訊及服務，已成為現今網路技術中相當重要的課題。「語意網」的架構主張將網路上文件內容以「知識本體」方式來將有意義的內容結構化，以建立起一個資訊明確及知識可重複使用的網際空間，進而促使網路上的資源及服務更易取得和分享。本研究嘗試從服務整合的觀點，設計旅遊語意網整體服務系統架構，目的是提供使用者便利、個人化、且幾近自動化的旅遊行程規劃與旅程服務。此雛型系統由使用者介面模組、旅遊服務排程模組、旅遊服務提供模組及旅遊服務註冊中心所組成；系統運作方式是由使用者提出旅遊安排的個人喜好需求及資源限制條件，接著旅遊服務排程模組依據需求，和不斷地透過資訊檢索的動作，使用綜合評估來篩選景點，並由旅遊行程規劃器進行行程安排。一旦使用者確認該行程規劃後，行程服務預約機制將會自動進行相關行程服務的預訂事宜。最後，系統即可將完整的旅遊行程規劃暨服務預約狀態，透過使用者介面模組呈現給使用者。本研究的主要貢獻計有：⑴設計暨建立台灣地區城市旅遊知識本體之雛型；⑵使用知識本體的樹狀結構，有效提昇檢索的效能及準確性；⑶將貪婪法理論、排列組合與綜合評估法則進行整合，以達到客製化的旅遊行程安排；⑷與服務預約機制相結合，藉以達到旅遊整體服務的功效。

The issues of effectively using Web information and services is getting more important among Internet technologies. The Semantic Web was thus proposed to facilitate resources on Internet re-usable and sharable using the ontology approach. In this paper, a novel prototype system, ”Travel Semantic Web of the Integrated Service System Architecture (TSWISS)”, was proposed to provide an integrated Web service for tourists making personalized travel plan more easily. TSWISS comprises four modules: user interface module, travel scheduler module, service provider module, and service registration center module. This research mainly focuses on three folds: the ways of defining and constructing Taiwan tour ontology, an efficient mechanism of information retrieval on Semantic Web, and a personalized trip scheduler. Relevant tour information was first collected to build domain and information ontology. Then, the constructed information ontology was translated into relation databases from original DAML format to make indexing more efficient. Furthermore, we utilized the top-down management of Web services to store the domain and information ontology in the registration center in advance. A novel information retrieval system was constructed using tag hierarchy within ontology and then was compared its performance with Google. Experimental cases were used to illustrate the results of the proposed system being more accurate and less redundant. Therefore, the novel system architecture based on Semantic Web could be expected to provide personalized travel plans for tourists more efficiently by using the designed ontology and proposed information retrieval mechanism.