Intelligent Information Retrieval based on Metadata
Abstract
Information retrieval has a long and successful history, but only the extraordinary information offering of the Internet has boosted new efforts to improve the quality of retrieved information by enriching both queries and data descriptions by semantics. This is also a valuable approach for other application areas, where the semantics of the data is not any more determined by a limited and well controlled set of applications, like geographical databases, knowledge management systems, or integrated databases for virtual enterprises. The use of formal ontologies as structured metadata in order to describe semantics appears to be a promising first step.
In this talk I shall report on the progress achieved in the TZI on the theoretical background of combining distributed ontologies at the semantical level, and on the (semi-)automatic generation of metadata including spatial representations for structured and unstructured information, also for multimedia. Among others, examples will be presented from the areas of geographic information systems and semantic video analysis.
|
Dr. Herzog received his M.Sc.in Applied Mathematics from the University of Bonn, Germany, and his Ph.D. in Computer Science from the University of Dortmund, Germany, in 1972 and 1976 respectively. From 1977 to 1993 he held numerous technical and managerial R+D positions in IBM Germany. Since 1993 he holds the Artificial Intelligence chair in the Mathematics and Computer Science Department of the University of Bremen, Germany, where he also initiated the Center for Computing Technologies in 1996, a software research and technology transfer center with more than 100 full- time researchers. His research interests include semantic-based analysis and retrieval methods for multimedia, multiagent systems in logistics, data mining in gene analysis, and wearable computing applications. |