Nested Hierarchical Dirichlet Processes
This addresses the need for more adaptable hierarchical topic modeling in large-scale text analysis, though it is an incremental improvement over existing methods like the nCRP.
The paper tackles the problem of rigid topic modeling in hierarchical structures by developing a nested hierarchical Dirichlet process (nHDP) that allows flexible word-level paths to topics, demonstrating it on datasets of 1.8 million and 3.3 million documents.
We develop a nested hierarchical Dirichlet process (nHDP) for hierarchical topic modeling. The nHDP is a generalization of the nested Chinese restaurant process (nCRP) that allows each word to follow its own path to a topic node according to a document-specific distribution on a shared tree. This alleviates the rigid, single-path formulation of the nCRP, allowing a document to more easily express thematic borrowings as a random effect. We derive a stochastic variational inference algorithm for the model, in addition to a greedy subtree selection method for each document, which allows for efficient inference using massive collections of text documents. We demonstrate our algorithm on 1.8 million documents from The New York Times and 3.3 million documents from Wikipedia.