LDA

LDA groups documents by hidden topics, like sports, food, or science.

What it does

LDA groups documents by hidden topics, like sports, food, or science.

Topic Modeling: Finding themes • Latent: Hidden inside data

How it works

• Look at words in many documents.
• Guess which words often belong together.
• Show each document as a mix of topics.

Uses and applications

• A news app can group articles by topic.
• A library can organize documents automatically.
• A teacher can sort student feedback into common themes.

In short

LDA finds topics in groups of documents.

Related methods

• TF-IDF: Also works from word counts, but scores words rather than finding topics.
• K-Means: Groups items into clusters instead of mixing topics in each document.
• Word2Vec: Represents word meaning as vectors instead of topics.

Back to Language and text models · All topics

The interactive demonstration and the Python walkthrough for this topic need JavaScript.