there is no dedicated node to do that. I would try to do pos tagging and filter out all terms but the verbs. Then try to identify the ending of the verbs ("ed" for past tense, "ing" for infinitive and so on) by using regular expressions. This would work of course only for regular verbs. For irregular you would need a dictionary.
Thanks for your response and sorry to revert back to this issue late / again.
Whereas I could carry out what you have suggested using the POS tagging and filtering, I would appreciate if you could provide some pointers for building the dictionary that you have referred. Are there ready dictionary list of words? Has anyone implemented this combination that I could benefit from. My goal is to predict at least the simple tenses viz. past, present and future with reasonably high degree of accuracy.
there is no dedicated node to apply JAPE rules on documents. There is also no node for verb or noun phrase chunking / tagging. You would need to implement that yourself. However, implementation of a tagger node is not that hard if you know how to code Java. Here is a tutorial: http://tech.knime.org/for-developers-integration-of-custom-tagger