Latent Dirichlet Allocation. Step by step tutorial?

Question

Algorithms

I would like to find suitable articles that explain not only the model itself, its principles, but also show how and what to do, in what order, where to transfer which data, how to use the results. Most sources miss the right information, and without it, the work stops. By the necessary information, I mean where to get the alpha, beta variables, themes, dictionaries, the order of data processing, the interpretation of the results. If all the same my guesses are correct and the topics and words of the topics are calculated from the input data during the operation of the algorithm, then I would like to find a resource that explains what to take, where to transfer, how to use the result.
PS saw Infer.Net - not quite right. This is a finished implementation, but I would first imagine all this in my head.

Answer 1

I was looking for that too, but I didn't find it. LDA is similar to plsa, it might be a good place to start if you haven't tried it yet. They write that there the accuracy is slightly lower, but you can try. Pretty clear written plsa in C https://github.com/JFantasy/plsa. It is easily rewritten in c#.