An efficient algorithm for topic ranking and modeling topic evolution
DEXA'11 Proceedings of the 22nd international conference on Database and expert systems applications - Volume Part I
| Hi-index | 0.00 |
Topic summarization and analysis is very important to understand an academic document collection and is very paramount for scientific research, which can help researchers find the hot field. Many scholars used the topic model to analyze the theme development, such as LDA. However, these methods need a pre-specified number of latent topics and manual topic labeling, which is usually difficult for people. Aiming to this problem, this paper proposes a method to analyze theme development with topic clustering. Different from the existing works, this paper uses the sliding window to cluster topics extracted in different time incrementally, the topic distance can be measured with KL-divergence. Some experiments on real data sets validate the effectiveness of our proposed method.