Temporal changes of knowledge in a distributed information system

In this article a method was described for knowledge changes analysis in a distributed information system. The source of knowledge is data collected and processed in separate subsystems. The data can be correct or incorrect, and consequently, the knowledge collected in the subsystems is consistent or inconsistent with the reality it describes. To recognize inconsistent, erroneous or false knowledge a semantics analysis is necessary. In this paper we assume that the knowledge collected in all subsystems correctly describes reality. We are interested in the changes of that knowledge in the system. The bases of the analysis are conspicuous states of knowledge and the distance function with expresses the difference between states. In this paper we defined temporal changes of the knowledge and described a method to measure temporal changes based on the analysis of the knowledge states and a method to measure these changes based on the analysis of the representation of the states of knowledge collected in the subsystems. Generally, the measurement of the temporal changes on the basis of representation is fraught with errors. The representation error is represented with the help of an information loss function. We described the properties of the dynamics analysis methods. We also added examples of the method applied in three different fields.