Semantic Web: Graphs, Imprecision and Knowledge Generation

Abstract

Growing interests in Artificial Intelligence generate expectations for constructing intelligent systems supporting humans in a variety of activities. Such systems would require the ability to analyze data and information, as well as to synthesize new elements of knowledge. To make it possible, there is a need to develop techniques and algorithms capable of extracting logic structures from data, processing and modifying these structures, and creating new ones. In this context, a graph-based representation of data—in particular a knowledge graph proposed by the Semantic Web initiative—is of a special interest. Graphs enable representing and defining semantics of data via utilization of multiple types of relations. Additionally, imperfections associated with processes of collecting and aggregating data and the imprecise nature of knowledge in variety of domains require data representation formats that can handle vagueness and uncertainty. In our opinion, knowledge graphs combined with elements of fuzzy set theory represent important building blocks of knowledge generation procedures. This chapter contains a brief overview of a framework enabling construction of systems capable of extracting logic structures from data and synthesizing knowledge. Utilization of fuzziness enables representation of imprecision and its inclusion in knowledge generation processes.