Combing LDA and Word Embeddings for topic modeling
Published in
3 min readSep 15, 2018
Latent Dirichlet Allocation (LDA) is a classical way to do topic modeling. Topic modeling is unsupervised learning and the goal is to group different documents to the same “topic”.
A typical example is clustering news to the corresponding categories including “Finance”, “Travel”, “Sport” etc. Before word embeddings, we may use Bag-of-Words most of the time. However, the world…