How do I convert topics to words?

[PaulDMendoza]

The example is missing the step where the topics identified are converted to a high level word that describes the topics for each document. Can you add that bit to the example?

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 87b08415-641d-4246-280f-b19e3baa8857
• Version Independent ID: 30d278f5-27f7-eee4-141f-d05dc1d34bfa
• Content: TextCatalog.LatentDirichletAllocation(TransformsCatalog+TextTransforms, String, String, Int32, Single, Single, Int32, Int32, Int32, Int32, Int32, Int32, Int32, Boolean) Method (Microsoft.ML)
• Content Source: dotnet/xml/Microsoft.ML/TextCatalog.xml
• Product: dotnet-ml-api
• GitHub Login: @sfilipi
• Microsoft Alias: johalex

[codemzs]

@justinormont can you please comment and close on this one?

[codemzs]

Hi @PaulDMendoza Thanks for the suggestion but this is a sample for LDA Transform and goal here is just to show how to use this particular API and not really how to do data science.

[PaulDMendoza]

@codemzs But 99% of people using this sample are going to have the same question. It's like you told a story but left off the ending.

Is it more than a couple lines to get the topic names or even the words that make up each topic and their frequency? I've looked at this API a bunch and can't figure it out. It seems like the LDA Transform should output a list of words for each topic but I find the ML.NET API very confusing for how to get to that data.

[nukeandserve]

I have the same question. How can I see the summary terms for a topic. I have the default 10 summary terms per topic set, but after fitting I cannot find the terms.

[justinormont]

See more detail on LDA topics in #4322