How to identify themes?

Amna
Amna New Altair Community Member
edited November 5 in Community Q&A
I am actually working on a study where I have collected open-ended responses from the respondents about their views related to brand activism. I have to identify themes from the data in terms of how many of the respondents (%age) think that brands remain quiet during certain social movements because they want to remain apolitical etc. I started trying it on Rapidminer but I am unable to get the desired results as I am new to the software.
Tagged:

Answers

  • Caperez
    Caperez Altair Community Member
    Hi @Amna, 
    Can you share more information about your dataset and process in Rapidminer?
    Why you consider that you aren't having the desired results?
    where is the problem?

    Best
  • Amna
    Amna New Altair Community Member
    The dataset is like customer reviews. It is in textual form. I want to identify themes out of those responses/reviews. For instance, a certain percentage of customer Brand is loyal, another %age of responses show that Brand is socially active etc. Which technique should I use to come up with this kind of output? I think it's going to be a combination of clustering and word association or probably LDA.  
  • Amna
    Amna New Altair Community Member
     or something like attribute extraction techniques and classifiers.

Categories