How to "chunk" based on field content

I am fairly new to KNIME, so apologies if this is “obvious”. Was not able to find any references thus far:

I am trying to implement a simple text processing pipeline, leveraging the sample workflow “19_Analyse_and_Visualize_Job_Postings”. As input, I replaced the table reader with a Tika parser. So far so good.

However, at the end of the pipeline, I would like to create one tag cloud per document, instead of one tag cloud over the entire set.

The last node before the tag cloud node, a Frequency Filter, contains each a row for each relevant term and it’s frequency by document. What I am now trying to do is instead of displaying a tag cloud for the entire output of the Frequency Filter, create tag clouds by document. I looked at chunking, but that seems to have only number of rows as a parameter. Equally, group-by also does not seem to work, as it changes the structure of the data.

Any help how to achieve this would be greatly appreciated!

1 Like

Look at


Hi @petervk welcome to KNIME forum

As @izaychik63 suggested the Group Loop is the way to go. When you follow the workflow below you collect every tag-cloud in the same table
Screenshot from 2020-03-06 20-05-21


Thanks. This was the clue I was looking for!

1 Like

Thank you. It all starts making sense now!


This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.