Definition
A type of statistical model used to automatically discover the abstract "topics" that occur and are distributed throughout a collection of documents.
In short: A type of statistical model used to automatically discover the abstract "topics" that occur and are distributed throughout a collection of documents.