LDA-based model for topic evolution mining on text
- Resource Type
- Conference
- Authors
- Wu, Qingqiang; Deng, Xiang; Zhang, Caidong; Jiang, Changlong
- Source
- 2011 6th International Conference on Computer Science & Education (ICCSE) Computer Science & Education (ICCSE), 2011 6th International Conference on. :946-949 Aug, 2011
- Subject
- General Topics for Engineers
Computing and Processing
Markov processes
Educational institutions
Data models
Evolution (biology)
Probability distribution
Mathematical model
Computational modeling
topic model
Latent Dirichlet Allocation
evolution
Gibbs sampling
- Language
A text mining model for topical evolutionary analysis was proposed through a text latent semantic analysis process on textual data. Analyzing topic evolution through tracking the topic different trends over time. Using the LDA model for the corpus and text to get the topics, and then using Clarity algorithm to measure the similarity of topics in order to identify topic mutation and discover the topic hidden in the text. Experiments show that the proposed model can discover meaningful topical evolution.