Towards intelligent agent mediate knowledge base integration

• Main Authors: Duen-Kai Chen, 陳惇凱
• Other Authors: Ruey-Shun Chen
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/62188643312367420240

博士 === 國立交通大學 === 資訊管理研究所 === 95 === Creating, retrieving, distributing and analyzing heterogeneous information poses significant challenges for future decision-making. With the rapid development of technologies for high-throughput data production and the potential for enormous distributed databases, it is becoming increasingly clear that sophisticated methodology and software solutions will be required to harness the potential of these new vast sources of information. In this study, a framework for integrating data source and software tools by ontology-driven Multi-agent systems approach and Web service technology has been developed. Software services for dynamically retrieving and manipulating information from various knowledge base and platforms in this framework are wrapped as Web Services that are described semantically in terms of domain ontology. The Web Services are registered in a brokering service so that it can be dynamically discovered and invoked by personal agent, which represent user, through matching queries against their semantic descriptions. Two illustration examples implemented in JADE agent platform are provided in this study in order to explain and validate the proposed framework. First example regards to construct Virtual Observatory; while the second example demonstrates how federated search in electronic library can benefit from the proposed framework The contributions of this dissertation can be summarized as follows. Users may invoke a set of services to accomplish the tasks to be done with the assist of personal agent. Due to the autonomous characteristic of agent, it is believed that the users may delegate information processing to personal agent and focus on their core competence. The capabilities offered by agents to query and invoke semantically enriched Web services is based on enhanced registries enriched with semantic information that provide semantic matching to service queries and published service descriptions. Furthermore, new services, agents, and users can be easily added to the proposed framework, thus provide an extendable and flexible open system.