Key Word analysis

[EdEvetts]

Hi

I have a data model with some columns which hold text responses, from which I would like to extract all of the words individually into a list for use in a report. I know something like this can be done using key phrase detection in Cognitive Services, but does anyone know of a way this can be done inside of Power BI?

thanks Ed

[v-ljerr-msft]

Hi @EdEvetts,

Could you be more precisely with your requirement by posting some sample data and your expected result, so that we can better assist on it? 

Regards

[dkay84_PowerBI]

Do you literally want EVERY word?  Or just key words/topics/entities (i.e. a machine learning solution)?  Or do you have a list of stop words that you want to omit?

[Greg_Deckler]

Can you post a bit of sample data and what you would like to end up with? For example, let's say you have this:

ID,Text

1, These are some words

2, These are some more words

And you want:

ID,Word

1,These

1,are

1,some

1,words

2,These

2,are

2,some

2,more

2,words

?

[EdEvetts]

Hi @Greg_Deckler @v-ljerr-msft @dkay84_PowerBI

Thanks for getting back to me and apologies for the vagueness. Just to give my question a bit of context. I have a Risk report and I am using the word cloud visual in the report to pick out the top 30 words from risk commentary - you can configure the word cloud with stop words. When I use the cloud to filter the risks in a table on the report - based on the word selected - it works but only temporarily. Once you click anywhere else on the report it loses the selection. I did hope I could use the include function available in some charts, but couldn't get that to work. So I was thinking I could create a list of the top 30 words displayed and then use this to filter the data. 

The data looks like this 

Risk Number	Risk Description
1	Extend business intelligence to everyone. Get dedicated capacity for your team and enhanced performance—without purchasing per-user licenses.
2	Go from data to insights in minutes. Any data, any way, anywhere. And all in one view.
3	Power BI is a suite of business analytics tools that deliver insights throughout your organization. Connect to hundreds of data sources, simplify data prep, and drive ad hoc analysis. Produce beautiful reports, then publish them for your organization

Into this (not an exact replica, but hopefully demonstrates the point), from which I could then create a slicer or something like that.

Risk Number	Key Words
1	business intelligence
1	capacity
1	enhanced
1	performance
1	purchasing
1	per-user
1	licenses
2	data
2	insights
2	minutes
2	data
2	anywhere
3	view
3	Power BI
3	suite
3	business analytics
3	tools
3	Deliver
3	insights

thanks Ed

[dkay84_PowerBI]

Since your key words include individual words and n-grams, you are really looking at something more like topic or entity extraction.  You can configure a machine learning model or use something like Cognitive Services from MSFT to do this type of work, and then load the results into Power BI to give you the results you are looking for.  I don't think just using a list of stop words will allow you to acheive your desired result.

[EdEvetts]

OK thanks @dkay84_PowerBI I'll try that route

thanks Ed

[Vvelarde]

@EdEvetts

A temporary Solution will be Work with Edit Interactions to prevent when you click in another chart lose the selection in WordCloud Visual.