Research On Concept-Cognitive Learning Based On Formal Context Of Network

In data analysis,concept-cognitive learning from the network is an important issue in the field of machine learning and artificial intelligence in the context of network.In previous research,people found that many data have their network structure and concept context,so combining the two to study concept-cognitive learning based on the formal context of network has important theoretical significance and application value.Firstly,this paper puts forward the concept of formal context of network,and combines the matrix reflecting the network structure in the analysis of complex networks with the formal context,and a formal context of network is proposed.In this way,the research of complex network analysis and formal concept analysis can be unified into a framework.It is a bridge between the above two theories,so that the research will be more deeply and meaningful by combining their respective advantages.Secondly,based on the formal context of network(NFC1),the following contents are studied from the micro,meso and macro hierarchies of network concept-cognitive learning:(1)The micro-hierarchy of network concept-cognitive learning.By defining the structural influence and connotation influence of the nodes and weighting them,the network influence of the nodes is defined.It combines the analysis of complex networks and the information of formal context to more finely describe the network characteristics of network micro concepts(individuals).(2)The meso-hierarchy of network concept-cognitive learning.By defining the network concept,the eigenvalues of network concept: concept potential,concept average degree are given.Not only can we dig out the network concept,but also describe the importance and local potential difference of the network concept in the network through the two eigenvalues of average degree and average potential.It also defines strong concepts and weak concepts,which can be used to analyze the diffusion and contraction of the network.(3)The macro-hierarchy of network concept-cognitive learning.The related algebraic properties of the strong(weak)concept are studied,which provides a theoretical basis for constructing the corresponding algebraic systems and generating various network concept spaces.Finally,the definition of the direction and power multi-relational formal context of network(NFC8)is given.The definitions and algorithms of NFC8 nodes influence and network concepts and their eigenvalues are studied.The theoretical significance of the above-mentioned theory and research on the study of the combination of complex networks and formal context is cleared through two examples.