How to set some words as 'unmodifiable'?

Hi
I have heaps of tweets on climate change and I’m learning sentiment analysis.
In my tweets are specialist expressions like ‘fridays4future’, ‘schoolstrike4future’ and equally ‘fridays for future’ or ‘school strike for future’. I’m thinking I need to convert ‘fridays for future’ various versions to ‘fridays4future’ first. Would rule engine be best for this?

How do I set ‘fridays4future’ etc UNMODIFIABLE? i.e. I’d like these established terms to remain complete with the number 4 in the middle etc. and appear in the final sentiment word/tag clouds.

I aspire to sentiment tag clouds (posit vs negative).
and did I understand correctly, that the sentiment for each ‘fridays4future’-term will be determined by the whole sentence sentiment? i.e. ‘fridays4future’ could actually feature in both positive and negative word cloud?

This is my first solo-project, trying to figure out Knime logic.
Many thanks for all your help!
Outi

Hi @outiloni,

I’d recommend to use the Dictionary Tagger to do this. It requires an additional input, e.g. a list of words/terms that you want to tag. While tagging the words will be set to unmodifiable. Afterwards you can either filter the documents for the tags using a Tag Filter or filter for unmodifiable terms using the Modifiable Term Filter.

Best,
Julian

2 Likes

Great idea, thanks Julian!