The business problems I faced was that after several meetings on the same topic, multiple minute of meetings were written, and members were then trying to summarize them manually. Can't there be a way of analyzing these automatically and getting suggestion for common themes?
Searching the web for open source tools for text analysis, I found Orange which is a very easy and robust tool for data mining.
Further inspired by the blog by Ajda Pretnar Žagar, a visualization for LDA topic modelling was done on business documents.
For this repository, the topic modelling was done on a sample text found on the web on Fundamentals of Business - Canadian Edition which is freely available from OpenEd under CC BY-NC-SA 4.0. Key takeaway were chosen from four randomly chosen chapters:
- "Ethics and Social Responsibility"
- "Foundations of Business"
- "Entrepreneurship: Starting a Business"
- "Structuring Organizations"
The excerpts are available in Fundamentals_of_Business.tab.
Widgets were configured as per the blog: 
In "Topic Modelling" I choose four topics, and we can for example see "Topic 3" seems related to ethics: 
"Topic 4" seems related to organizational structure: 
The topics of foundation and entrepreneurship however were not identified based on this sample.
The analysis is available in lda_analysis.ows and should work with Orange version 3.31.1.