The Exploration Of Children's Self-Concept Structure

Self-concept is an individual's perception of all aspects of himself,and its structure is multi-dimensional,multi-layered and organized.Self-concept is an important component of personality,which reflects the individual's ability,appearance,social acceptability,attitude and emotion.This paper is to explore children's self-concept structure by analyzing children's open self-introduction text with the technical method of knowledge mapping.In addition,the development of natural language processing technology makes it possible for knowledge mapping to display self-concept structure.In this study,we established a map of children's self-concept knowledge,and applied the map to interest discovery,information association,relational classification query and other aspects.In the process of constructing the knowledge map,the bi-lstm and CRF fusion methods were used in this paper to extract the entities of the names of people and organizations.The extraction accuracy of this method was higher than 80%.Later,this study USES the dependency grammar-based method to realize the open joint extraction of entity relations,and extracts a large number of(entity,relationship,entity)triples as a supplement to the former.After that,the text classification is realized by training the convolutional neural network and the text is divided into the self-concept structure framework of this paper.The classification accuracy of convolutional neural network is more than 80%,and the possibility that the text belongs to the classification is output in the classification prediction,so as to serve as a reference for coding correction.Finally,this paper proposes a storage method based on Neo4 j graph database to visualize children's information and self-concept structure reflected in their selfintroduction.Based on the analysis of open text,this paper concludes that(1)there are differences in the self-description of children in different educational stages,the components of children's self-concept are affected by the education stage,and the secondary school group can describe themselves in more aspects than the primary school group.(2)in the study of the open self-introduction text,this study found that children construct themselves from multiple perspectives and evaluate themselves from various aspects,so this paper proposed a two-dimensional network structure of children's self-concept.This paper holds that children's self-concept is based on five perspectives: fact,will,self-evaluation,sociality and tendency,and a certain selfconcept index may have multiple contents.(3)this study provides a reference and direction for the scale revision and scale compilation.Study found that more children has strong values in social,academic,past through children's self-concept scale research on various aspects such as weight,but in the open text analysis of this research is not the case,primary and secondary school children are very care about their academic performance and social development,and the Angle of the researchers need to carefully consider the ego concept of index weight,more need to consider some specific indicators for Chinese children the importance of the degree and the significance.This research has a new research form and direction.It not only puts forward Suggestions for future related research,but also has a wide range of final results in practice.The map can be popularized and applied in junior and senior high schools as a tool of school information management,which can better serve schools in carrying out educational activities and research-based learning.