Text analysis using Alteryx

[fangyi189]

Hi all, I am trying to do out a task where I'll extract out each keyword's occurrences, and count them to see the top 10 keywords. Any ways to do it out?

[andre347 · Accepted solution]

I would split my text on a \s (whitespace) with the text to columns tool (split to rows). Then make every single word lowercase (or upper case). Use the Regex tool or the Data Cleansing tool to remove unwanted characters. Summarize on word count and then use a sort descending on this field. After that you can use a sample for the top 10.

[fangyi189]

Sorry, do you have a workflow to show it out?

[andre347]

Sure. See attached. Created in Alteryx v11.7

[fangyi189]

Hi, so thankful for your help (: thanks

[mayscw10]

We have done some similar analysis. Another cool view is to use a Multi-Row Formula once you have the keywords to look for common word pairings. Due to my organizations policies I can't upload my workflows but the expression to combine the words is "[Keywords]+"~"+[Row+1:Keywords]" this will add a tilde ("~") instead of a space between the Keywords.