AI in Dataflow, Power BI webservice, Cognitive Service -Part1

These days, Dataflow (Power Query in Power BI Service) is a new feature that has been released. according to[1]:
Dataflow is a Power Query process that runs in the cloud independently from any Power BI reports.

If you want to know more about dataflow see [1]


There is a new feature in dataflow that allows users to use the AI in Power BI Service! in this series, I am going to show you how you able to use this feature.

Part one, you will see how to use Ai insight for the aim of Text Analytics 

Part two- Three, How to use the embed machine learning algorithm in Power Query

Part three or four: How to bring your own machine learning algorithm using Azure Machine learning workspace in Power BI service


In this Part, we are going to see how Text Analytics feature is available in Power BI service dataflow.



Some configuration note

You need to have a premium account, However, currently it is in private preview, based on what I found in  their  website:

“These features will be available for a public preview by March 2019” 

which is in 2 months!


I requested for private preview couple of months ago (because of lots of interests it is not available anymore) 

In this post, I will show how you able to use this awesome feature.


First, you need to login to your Power BI Service,

In another workspace (not My workspace) you able to create dataflow.

to create a new workspace, click on the Workspaces-> Create app workspace

Now, in the new page,  you will see a page that able us to create a new workspace (pro or premium)

To do that, click on the Try Now.


A new page will be open, that allows you to create a new workspace with image, name and description

choose proper image, name, and description and click on the Save.

In the new page, you able to see your new workspace and the welcome page, just skip the welcome page to navigate to the main workspace page.


In the main workspace page,  beside dataset, Dashboard, Reports, we have new feature name as Dataflows, click on Create option at the top right the page and choose the Dataflow.

it navigates to a page that asking you about adding entities (data source) click on the define new entities.

In the next page, you able to connect to different resources from cloud to on-premises. For this example, we just put some manual data to check the cognitive service (Text Analytics),


Click on the Blank Table, then change the title of the column to Comments,  put the below text for each row ( row by row). then put a name for the table and click on the Next.

I ordered three widgets last week and just received them. I am VERY disappointed in the way they were packed. The box was only about two thirds full of packing peanuts and the widgets were not individually wrapped. As a result, they were able to bump up against each other while in transit. The display on one doesn’t work at all and the other two are quite scratched up (look like they have been in service for years) though they do work fine. I have always been satisfied with your service and trust you will make this right, but I suggest you look into how you are packing these items. You will have a lot of unhappy customers otherwise.
Can’t believe you fools shipped me the wrong items AGAIN. If you weren’t the only company I’ve been able to find that even has these items, I would have stopped ordering from you a long time ago. I order weekly and it seems like at least once a month my order has some kind of foul-up. Waiting for the order to be corrected would set my business back so I need to keep extra on hand just in case. Not fair for me to have to carry extra inventory to make up for YOUR screw-ups! Hire someone who can tell the difference between a thingamabob and a whatchamacallit. If anyone reading this knows of another source please let me know.
Wow! I had NO IDEA that reciprocating garombles of such high quality even existed! And they’re a steal, too; I’ve paid twice your asking price for garombles that weren’t nearly as smooth in operation. Free ground delivery was a plus (the overnight shipping seems a little spendy though). We’ll see how these actually hold up, but so far I’m sold!
That hand model needs a manicure, stat!

the new query will be generated and navigate into a new page as Power Query editor in Power BI web service. As you can see in the below picture, It has some of the features of the Power Query Editor from transforming the columns, to combine and so forth.

For the aim of text analytics, there is an icon on the top of the page as AI Insight. click on it

by clicking on the AI insight, it asks you about the credential and authentication for login to account.

then click on connect.

then a new page will show up, in new page cognitive services will show up: there are four different cognitive services are available. click on the last one that is Sentiment Analysis

Sentiment Analysis is for identifying how much a customer is happy with os or not. the value is something between 0 to 1. closer to 1 means customer happy close to 0 means omer is not happy.

Choose the service. , it asked about the Text, that you need to mention which column you want to apply this service. in this example is Comment is the target column.

Then, just click on the Invoke. A new page will be shown that shows the original column (Comment) plus a new column added name Cognitive Service, this new column has a value from 0 to 1 for each row.

Now we got the result, we need to save the result so we able to use it in power BI service report or later on in Power BI desktop. Click on the Done at the bottom of the page to apply all changes, then, you need to save the query by clicking on the top right of the page and put a name for your query.

Then, Under the Dataflow you able to see your dataflow. You just need to click on the Refresh Botton.

In this example, first I am going to get the data in Power BI desktop.

To do that, Open your Power BI desktop, and sign in to account you are using for the Dataflow. Click on Get Data choose the Power BI dataflow. It will ask about the sign in.


After sign into the account, now you able to see the list of dataflow you have. Expand it and see an overview of the query with Comments and sentiment score columns.

still you not able to create a report using dataflow in Power bI service.

In the next post, I will show you how you able to classification and prediction using some other services in Dataflow


Leila Etaati on LinkedinLeila Etaati on TwitterLeila Etaati on Youtube
Leila Etaati
Trainer, Consultant, Mentor
Leila is the first Microsoft AI MVP in New Zealand and Australia, She has Ph.D. in Information System from the University Of Auckland. She is the Co-director and data scientist in RADACAD Company with more than 100 clients in around the world. She is the co-organizer of Microsoft Business Intelligence and Power BI Use group (meetup) in Auckland with more than 1200 members, She is the co-organizer of three main conferences in Auckland: SQL Saturday Auckland (2015 till now) with more than 400 registrations, Difinity (2017 till now) with more than 200 registrations and Global AI Bootcamp 2018. She is a Data Scientist, BI Consultant, Trainer, and Speaker. She is a well-known International Speakers to many conferences such as Microsoft ignite, SQL pass, Data Platform Summit, SQL Saturday, Power BI world Tour and so forth in Europe, USA, Asia, Australia, and New Zealand. She has over ten years’ experience working with databases and software systems. She was involved in many large-scale projects for big-sized companies. She also AI and Data Platform Microsoft MVP. Leila is an active Technical Microsoft AI blogger for RADACAD.

3 thoughts on “AI in Dataflow, Power BI webservice, Cognitive Service -Part1

Leave a Reply