Abstract | ||
---|---|---|
With a few exceptions, extensions to latent Dirichlet allocation (LDA) have focused on the distribution over topics for each document. Much less attention has been given to the underlying structure of the topics themselves. As a result, most topic models generate topics independently from a single underlying distribution and require millions of parameters, in the form of multinomial distributions over the vocabulary. In this paper, we introduce the Shared Components Topic Model (SCTM), in which each topic is a normalized product of a smaller number of underlying component distributions. Our model learns these component distributions and the structure of how to combine subsets of them into topics. The SCTM can represent topics in a much more compact representation than LDA and achieves better perplexity with fewer parameters. |
Year | Venue | Keywords |
---|---|---|
2012 | HLT-NAACL | component distribution,compact representation,topic model,shared components topic model,fewer parameter,underlying structure,underlying component distribution,better perplexity,multinomial distribution,single underlying distribution |
Field | DocType | Citations |
Dynamic topic model,Perplexity,Latent Dirichlet allocation,Normalization (statistics),Computer science,Multinomial distribution,Natural language processing,Artificial intelligence,Topic model,Vocabulary,Machine learning | Conference | 8 |
PageRank | References | Authors |
0.49 | 13 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Matthew Gormley | 1 | 84 | 10.25 |
Mark Dredze | 2 | 3092 | 176.22 |
Benjamin Van Durme | 3 | 1268 | 92.32 |
Jason Eisner | 4 | 1825 | 173.00 |