Toward semantically enabled on-line tour planning: A conceptual framework and an ontology model
MetadataShow full item record
This dissertation presents a three-level conceptual framework and a three-component ontology model for semantically enabled on-line tour planning. The conceptual framework includes an abstract level, a representation level, and an implementation level. The abstract level identifies two types of participants (the traveler and the travel information providers), two types of information (internal information and external information), and two types of processes (implicit and explicit processes). Subsequently, at the representation level, the two types of participants are represented as two types of agents, the IAT and the IATIPs, along with information and processes associated with them through an agent-based design. These specifications guide implementation of semantically enabled on-line tour planning systems. The ontology model serves as a basis for the understanding and integration of the information involved in semantically enabled on-line tour planning. This model includes three components: an ontology for the traveler, an ontology for the travel information providers, and mapping between the two ontologies. Corresponding to the two types of participants, the first two components are used to represent their perspectives, respectively. These two ontologies consist of concepts and their relationships that the semantically enabled on-line tour planning system needs to describe the internal information about the traveler, and the external information about travel information providers, respectively, in order to make a tour plan. The method "ontology learning from text" is used to derive these two ontologies. The ontology for the traveler takes advantage of the existing research in choice models of tourist attractions and destinations. The ontology for the travel information providers includes both local ontologies and a central ontology. This dissertation focuses on the development of the central ontology, which integrates the heterogeneous local ontologies to describe tourist attractions in unified meanings. For the third component in the ontology model, this dissertation proposes a two-level approach to mapping the two ontologies. At the first level, conceptual mapping uses formal concept analysis to identify equivalent or overlapping concepts in the two ontologies. Subsequently, the property values of mapped concepts are evaluated to create a personalized tour plan that matches the preferences of a traveler with the information provided at travel websites. A hypothetical case study is used to demonstrate the conceptual framework and ontology mapping approach, and to illustrate their application. A prototype of the semantically enabled on-line tour planning system is also developed to discuss implementation issues. The software agent technology, semantic web technology, and spatial web services technology are applied in the prototype implementation. An interactive and intuitive graphic interface is also designed in the prototype to display the recommendation results and obtain users' feedback.