A Burst Topic Detection Model Based on HVCH Model and Burst_NBT Method

This paper proposes a burst topic detection model based on HVCH model and Burst_NBT method. The model overcomes the sparsity difficulty of microblog by establishing HVCH model, effectively detects topics of microblog. Then the model establishes Burst_NBT method to calculate the burst scores of the words and cluster the burst words of microblog document sets of each topic, and to score and rank the burst word clusters to detect the corresponding burst topics of each topic. The experiment results indicate that the proposed HVCHBurst_NBT model improve the evaluation indicators compared with the other two burst topic detection models.