|BUTE VIK Department of Measurement and Information Systems||Printable version|
|DMIS homepage - Research - Projects - Hybrid intelligent framework|
Development of hybrid intelligent framework
2001 - 2004
Local supervisor: Strausz György
Efficient evaluation of information provided by distributed sources in different representation formats are one of the major open problems of intelligent system research and development tasks nowadays. The main reasons behind the emerging need is the hidden capability of huge number of available information stores. The purpose project targets two important area of distributed information retrieval: a.) Research of hybrid intelligent methods The project focuses on the integration technics of neural network and expert system (rule based, frame based systems, semantic networks) approaches. Different models are known in this research area, but the effectiveness of this solutions has not reached the level of the requirements of complex systems. Therefore the research of application oriented models are in high interest. b.) Development of hybrid intelligent famework One of the main obstacles of both hybrid intelligent system development and research are the lack of an appropriate development environment supporting neural-expert hybrid model developments, model demonstrations and tests. The project also aims to design and implement hybrid intelligent system development support tools that could be used for different research tasks in this field.