The
Text Filter node enables you to focus on the terms and documents that
are most likely to enhance your model. For most of your text mining
projects, you will want to follow the Text Parsing node with the Text
Filter node. This way, you can eliminate extraneous information caused
by the presence of noise terms and other terms that are not pertinent
to your analysis. If your model would be improved by focusing on a
subset of the collection, then the Text Filter node can remove documents
that do not fit your criteria. The end result of the Text Filter node
is a compact, yet information rich, representation of your collection.
For example, you can put a Text Filter node after a Text Topic node
and filter for documents that contain specific topics.