A Social Network Analysis Method for "One Stop" Student Community
In order to conduct social network analysis on online student communities,this paper proposes an outlier detection method based on student social data,aiming to identify potential outliers in social networks.This paper designs a feature extraction layer that includes connection features,activity features,and centrality features,and constructs student feature vectors through comprehensive analysis.Subsequently,a multi-layer perceptron is used as the neural network layer to mine student features and ultimately output the probability of students being social isolators.Finally,experiments are conducted on the basis of the Friendster dataset,and the results show that the proposed method has high accuracy,precision and recall rate.
NETL
NSTL
万方数据
