Buchtitel: SW-KM EKAW 2002
Finding right information in the right moment is quite a tedious task on the Web. The main difficulty is the wrong context in which information is found and, therefore, part of the problem lies in the process of providing metadata about information resources on the web. In other words, only by searching for keywords, which are assigned to the content of an information resource, a machine agent cannot differ between various contexts of using these keywords, and some non-relevant resources will be retrieved as the result of the searching process. In this paper, we propose a method for improving the annotation of information resources - by forming conditional statements about keywords used in the describing content of a resource. This kind of annotation leads to the better precision of the information retrieval process - by expressing the context of searching in a more precise manner. Moreover, such more expressive form of describing content of information resources supports a more powerful knowledge sharing process enabling the discovery of new information by considering the combination of existing information resources. In that way, the system provides some answers which are not explicitly stated in the information repository. We present a knowledge management framework, which implements such type of annotation and give a small evaluation study. The framework is supported by Semantic Web technologies, which are based on the machine-understandable description of document content, enabling in that way the automation of the knowledge sharing process.