Topic modeling for BBC News

As discussed earlier, the goal of topic modeling is to identify patterns in a text corpus that correspond to document topics. In this example, we will use a dataset originating from BBC News. This dataset is one of the standard benchmarks in machine-learning research, and is available for non-commercial and research purposes.

The goal is to build a classifier that is able to assign a topic to an uncategorized document.

BBC dataset

In 2006, Greene and Cunningham collected the BBC dataset to study a particular document—Clustering challenge using support vector machines. The dataset consists of 2,225 documents from the BBC News website from 2004 to 2005, corresponding to the stories collected...