Word Cloud Mod Analysis of Individual Words in Text Field

[Lauera1]

I have a column with user entered comments so there is a lot of inconsistency between entries.  However, I would like to be able to see if there are common words used between comments.  I have seen other word clouds programs have the ability to choose 1-3 word commonalities amongst free text fields.  Is there a way to do this with the Spotfire Word Cloud Mod or is there a different visualization that would work better for this application?

Thanks in advance!

[Sweta Kotha]

Hi! It sounds like some natural language processing could help.
Below is a community link for the spotfire-dsml python library; specifically the nlp_preprocessing submodule would be helpful.

The 1-3 word commonalities are referred to as N-grams (unigrams, bigrams, trigrams, etc.). I attached a sfd file (this is also in the exchange component of spotfire-dsml) that can be directly uploaded into a Spotfire data function to get these N-gram features. Then, you'd use the Word cloud to visualize these features.

All you need to specify in the 'Edit Parameters' section is the text column name, ID column, and perhaps modify the N-gram size/range (if you want more than trigrams or 1-3 common words - this can be higher but the default is 3)
 

[Modeling] NLP Python Toolkit - Features.sfd