A study of social graph analysis for preschool education using face authentication: Pruning for centrality optimization in graph analysis

Abstract

Social graph analysis for preschool education is discussed in this paper. Our social graph describes relations between preschool children and preschool teachers. The preschool teachers wear eyeglasses equipped with a video camera, and they take a video in their childcare. The social graph is created by a face recognition and authentication systems that perform identification for children and preschool teachers from the video. Some centralities of our graph are computed, and analyze the preschool education. When we take a long video, our social graph may become a complete graph, so that values of the centralities for all children are equal; it is our technical issues. In this paper, we discuss graph pruning and optimize our centralities for the graph analysis. Our performances of graph pruning are shown in our experiments.