Modelling of Experienced-Based Data in Linked Data Environment

The Resource Description Framework (RDF) proposed as a part of Semantic Web becomes an important way of representing data and information on the web. Its intrinsic feature of high connectivity has a direct impact on the ability to express semantics of data, as well as development of methods for extraction of information and knowledge embedded in the RDF data. In this paper, we proposed a procedure to construct a model of data that has been experienced, i.e., discovered and collected, by software agents and systems. The process is based on processing previously constructed concepts of data. The model is represented as a hierarchy that contains definitions of concepts, their instances, as well as relations found in the data. Each concept definition is represented as a set of features. The features are relations between a given concept definition and definitions of other concepts. The proposed methodology allows for determining importance of features and their degrees of contribution to definitions. This reflects realistic, real-life connections and dependencies existing between generalized concepts.