The Research On Incremental Concept Tree Generation Method Based On Attribute Topology

Incremental concept cognitive learning mainly involves the process of dynamic big data knowledge.Moreover,as a potential direction of cognitive computing research,formal concept analysis can describe in detail the specific process of concept cognitive learning.As a new expression of formal context in formal concept,attribute topology can clearly show the relationship between the new data and the original data,thus reduce the complexity of the process of concept cognitive learning,which already has preliminary research progress in the area of dynamic context analysis.Based on above two points,this paper will put forward an incremental concept tree generation method based on attribute topology which is used for the process of dynamic computing of incremental concept cognition.Firstly,this paper will propose the collapse and structured expression of incremental formal context based on the combination of attribute topology and the incremental concept cognitive learning method.The collapse of incremental formal context is a method which is realized by introducing different types of new objects which will have differential influence on original topology,in order to reduce the complexity of new objects cognition process.The structural expression of incremental formal context is focused on the analysis of the relationship among original objects based on new objects,which provides basis for latter incremental concept tree generation method.Secondly,this paper uses the expression of attribute topology based on formal concept analysis and detailed analyze the cognitive process of different types of new objects from the view of cognition.Because concept tree is a form of concept cognition,this paper presents the process and result of concept cognition by the form of concept tree.For the new non-global objects,the position where the new object adds to the concept tree should be determined first,then the subtree of concept tree that can grow following the new object should be determined according to the relationship between the objects in the incrementally structured representation,finally the concept tree updates according to principle of concept.Finally,the key frame images extracted by short video is used as experimental data to verify the generation method of incremental concept tree proposed in this paper,and concept cognitive simulation result will also be got,based on the story plot development.According to the analysis of the experimental process and result,we will see that when the order of increments is different,the result of concept tree which we finally got will also be different.This is similar to the cognitive process of human being.The incremental concept tree generation method based on attribute topology proposed in this paper enriches the research of attribute topology in dynamic incremental concept cognitive computing.