The Development of the Metagraph Data and Knowledge Model Yuriy Gapanyuk Bauman Moscow State Technical University, 2-ya Baumanskaya ul., 5, Moscow, 105005, Russia Abstract Among the cognitive models in artificial intelligence, graph models of knowledge representation traditionally play an important role. Currently, models based on complex networks or complex graphs are attracting increased attention. One of the most developed models of this class is the metagraph model. Currently, one of the most developed modifications of the metagraph model is the annotated metagraph model. While the metavertices in this model are primarily intended to describe data and knowledge, the metaedges are more intended to describe processes. Thus, the metagraph model allows describing data, knowledge, and processes within a single model. In order to transform the metagraph model, the metagraph function agent and the metagraph rule agent are used. The metagraph agent allows generating one metagraph based on another (using open rules) or modify the metagraph (using closed rules). Keywords 1 complex network, complex graph, emergence, metagraph, metavertex, metaedge, metagraph agent. 1. Introduction Graph models of knowledge representation traditionally play an essential role among cognitive models in artificial intelligence. Currently, models based on complex networks or complex graphs are attracting increased attention. According to [1]: “a complex network is a graph (network) with non-trivial topological features – features that do not occur in simple networks such as lattices or random graphs but often occur in graphs modeling of real systems.” The terms “complex network” and “complex graph” are of ten used synonymously. According to [2]: “the term ‘complex network,’ or simply ‘network,’ often refers to real systems while the term ‘graph’ is generally considered as the mathematical representation of a network. … The differences lay mainly in size (smaller for the graph and bigger for the network) and in the parameters and the tools employed to analyze both structures.” The most significant discrepancies are caused by the term “complex” in relation to graph models. As a rule, the term “complex” is interpreted in two ways: Way I. Flat graphs (networks) of a huge dimension. Such networks may include millions or more vertices. The edges connecting the vertices can be non-directional or directional. Sometimes a multigraph model is used; in this case, two vertices can be connected not by one but by several edges. It is precisely such a model in the literature that is most often called the “complex network.” Studies of this model are carried out mainly by specialists in the field of mathematics. Researchers consider such parameters as the distribution of the number of links between vertices, the allocation of strongly connected subgraphs. Often, a quantitative metric is introduced for relationships, which is usually interpreted as the distance between the vertices. Dynamic models are actively investigated, in which vertices and edges are randomly added to an existing complex network. Such models are of interest in Russian Advances in Fuzzy Systems and Soft Computing: Selected Contributions to the 10th International Conference «Integrated Models and Soft Computing in Artificial Intelligence» (IMSC-2021), May 17–20, 2021, Kolomna, Russian Federation EMAIL: gapyu@bmstu.ru ORCID: XXXX-XXXX-XXXX-XXXX © 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (CEUR-WS.org)