KNIME Textprocessing version 2.9 or later and the Palladian community feature is required to load and execute this workflow.
The workflow starts with a URL to a NY Times rss news feed. The news feed is downloaded and parsed and transformed in DocumentCells.
Names of persons, organizations and locations are then recognized and the corresponding tags are assigned, in order to apply a coloring based on a tag type later on.
After filtering of all non-persons, -organizations, or –locations and transformation into a bag of words, colors are assigned and the terms are visualized via a Tag Cloud.
|Tag Cloud organizations, location and persons which have been recognize bei the OpenNLP named entity recognizer. Note that the tag cloud supports hiliting.|