Mixing Dirichlet Topic Models and Word Embeddings to Make lda2vec
This work provides a method for creating interpretable document representations for scientists, though it appears incremental as it builds on existing word embedding and topic modeling techniques.
The authors tackled the problem of combining word embeddings and topic models by introducing lda2vec, which jointly learns dense word vectors and sparse, interpretable document-level topic mixtures. The model achieved this through a non-negative simplex constraint, making it simple to implement in existing frameworks.
Distributed dense word vectors have been shown to be effective at capturing token-level semantic and syntactic regularities in language, while topic models can form interpretable representations over documents. In this work, we describe lda2vec, a model that learns dense word vectors jointly with Dirichlet-distributed latent document-level mixtures of topic vectors. In contrast to continuous dense document representations, this formulation produces sparse, interpretable document mixtures through a non-negative simplex constraint. Our method is simple to incorporate into existing automatic differentiation frameworks and allows for unsupervised document representations geared for use by scientists while simultaneously learning word vectors and the linear relationships between them.