We use the KNIME Simple Streaming nodes to do the first part of the text processing. See the first wrapped metanode. To enable streaming (for streaming enabled nodes) right-click 'configure' then switch to the 'Job Manager Selection' tab, and select 'Simple Streaming'. In this case the first wrapped metanode is already configured for streaming, and the second is not (the second contains no streamable nodes). The workflow reads textual data from a csv file and converts the strings into documents. The documents are then preprocessed, i.e. filtered and stemmed and transformed into numerical document vectors. All the preprocessing magic takes place in the Preprocessing meta node. After the document vectors have been created the sentiment class is extracted and a predictive model is built and scored.
This is a companion discussion topic for the original entry at https://kni.me/w/nmIfdmNGfg3gXrtD