User Naming Conventions Mapping Learning for Social Network Alignment

Abstract

The existing research on social network alignment using usernames is mainly based on the similarity between usernames calculated by different classifiers. However, if the number of available annotations and training time are limited and feature extraction is incomplete, the accuracy of social network alignment would have been be reduced. Based on the above, this paper proposes a BP neural network mapping for social network alignment (BSNA). The BP neural network is used to realize the mapping between two social network user name vectors, and the classification problem is transformed into a mapping problem between vectors. The experimental results on several social network data sets show that compared with the benchmark method, the social network alignment precision of the proposed model is improved by 4%, and the experiments with smaller training set ratio and less training time have higher precision and faster convergence than the benchmark method.