You can use Text Mining. There you can use some steps to eliminate unnecessary (stopwords) and to reduce redundant words (stemming). You have to be aware also of synonymous and how you will score your matching words. That being said, you can run a test. The problem is that you should be able to categorize the words of interest in segments before the test and which group of words are related to the segment of interest.

I don't know a tool that can do this for you for free. The closest of a segmentation with text mining I made was sentiment analysis with R but you can code your segmentation specs in R as well and try to find your answers.

Hope this brief explanation could help you.

Cheers

Dear Sanja,

Well normally customer clustering (a segmentation method) is used in data mining. Example clustering customer habits in markets (i.e. buying habits) and this is done using K-means algorithm in Matlab as an example.

What you need to know is based on what categories you need customers to be clustered (cluster centers).

In your case may be you need first to filter your data for irrelevant information.

See the following site for more details https://www.business-science.io/business/2016/08/07/CustomerSegmentationPt1.html

Dear Alexander Brasil,

thank you very much for the answer. It will be very useful.

Best regards,

Sanja

Dear M. Awad,

thank you very much for the opinion and the resources. I am glad to follow it.

Best regards,

Sanja