Creating rules for Topics dictionaries in Text iQ | XM Community
Question

Creating rules for Topics dictionaries in Text iQ

  • 21 July 2023
  • 5 replies
  • 64 views

Userlevel 3
Badge +6

I found this article in the help section but it’s not answering my question:
https://www.qualtrics.com/support/survey-platform/data-and-analysis-module/text-iq/topics-in-text-iq/

I want to create a rule for a topic where it finds all the words that start with “think”. Right now, the automatic lemmatization is finding “thinking” but not “thinkers”. Is there something else I can add to the query to expand what words it’s looking for?


5 replies

Userlevel 6
Badge +18

Hi @AnnaHar,

The easiest way to achieve this requirement is to include “thinkers” into your code frame/ Query sting, e.g. your query string should look something like this, “Think OR Thinking OR Thinker” to cover as much combinations as you could.

 

Hope this helps.
 

Userlevel 3
Badge +6

Thanks @SaurabhPujare_Ugam .

That’s a little disappointing because I was hoping for a more advanced lemmatization process but since you’re the only one to respond with a solution, I suppose that’s all there is.

Userlevel 6
Badge +18

Hey @AnnaHar,

Yeah, I believe the lemmatization could be subjective at times, ex., in above case, it was ideal to to tag “thinking”/”thinkers” under “think”, at the same time, we DO NOT want to tag “carpet” under “car”.

Hope this makes sense :)

Userlevel 3
Badge +6

For anyone else who stumbles across this, I learned yesterday from Qualtrics Support that the main issue here is that it reads “think” as a verb so it doesn’t find nouns with that base.

Userlevel 6
Badge +18

Thanks for sharing, @AnnaHar. It is indeed a great insight.

Leave a Reply