Gaussian Process Topic Models
This work addresses the need for more flexible topic modeling in natural language processing, though it appears incremental as it builds on Correlated Topic Models with Gaussian Process extensions.
The authors tackled the problem of topic modeling by introducing Gaussian Process Topic Models (GPTMs), which incorporate document kernels and topic correlations, achieving improved performance in both topic quality and document embedding compared to existing methods.
We introduce Gaussian Process Topic Models (GPTMs), a new family of topic models which can leverage a kernel among documents while extracting correlated topics. GPTMs can be considered a systematic generalization of the Correlated Topic Models (CTMs) using ideas from Gaussian Process (GP) based embedding. Since GPTMs work with both a topic covariance matrix and a document kernel matrix, learning GPTMs involves a novel component-solving a suitable Sylvester equation capturing both topic and document dependencies. The efficacy of GPTMs is demonstrated with experiments evaluating the quality of both topic modeling and embedding.