This node trains a Doc2Vec (Paragraph Vectors) model on labelled documents. The model will learn sequence representations (vector for each label) as well as word representations (vector for each word), that can be extracted using the Vocabulary Extractor Node. For more information on Word Vectors in general see: http://deeplearning4j.org/word2vec
The KNIME Deeplearning4J Integration has been marked as legacy with KNIME Analytics Platform 5.0 and will be deprecated in a future version. If you are using this extension in a production workflow, consider switching to one of the other deep learning integrations available in KNIME Analytics Platform.