摘要
In this paper, a method for telecommunications fraud detection is proposed. The method is based on the user profiling by employing the Latent Dirichlet Allocation (LDA). The detection of fraudulent behavior is achieved with a threshold-type classification algorithm, allocating the telecommunication accounts into one of two classes: fraudulent account and non-fraudulent account. The accounts are classified with use of the Kullback-Leibler divergence (KL-divergence). Therefore, we also introduce four methods for approximating the KL-divergence between two LDAs. Finally, the results of experimental study on KL-divergence approximation and fraud detection in telecommunications are reported.