Click here to Skip to main content
14,034,985 members
Click here to Skip to main content
Add your own
alternative version

Stats

1.2K views
1 bookmarked
Posted 22 Mar 2019
Licenced CPOL

MebhiChokiDAR – Tweet Sentiment Analysis and More Using Microsoft Flow

, 22 Mar 2019
Rate this:
Please Sign up or sign in to vote.
How to do sentiment analysis using few clicks and see live graphical representation using Power BI live feed

Microsoft Flow provides various templates to achieve your goal and Twitter Sentiment analysis is one of them.

In this tutorial, we will see how to do sentiment analysis using few clicks and see live graphical representation using Power BI live feed.

As in India currently, #Mebhichokidar hash tag is very viral. So,we are using this hashtag to cross check how it is going.

Prerequisites

There are 3 basic prerequisites to complete this task:

  1. You should have Microsoft flow account. If you don’t have that, then you can create via link https://flow.microsoft.com
  2. You should have Microsoft Power BI account. If you don’t have, then you can create via link https://powerbi.microsoft.com/en-us/
  3. You should have Microsoft Azure Portal account for cognitive services. If you don’t have, then don’t worry, you can create via link https://Portal.azure.com

Once you have all the 3 accounts available, then proceed step by step as suggested below.

Step 1

Just open the Microsoft flow and select the template “Run Sentiment Analysis on tweets and push Result to a Power BI dataset” as shown in the below figure:

Indiandotnet_Twitter_1

Step 2

When you select the option, you will find the below screen as shown below:

Indiandotnet_Twitter_2

Step 3

We have to configure 3 connections as expected, text analytics connection (Azure cognitive service), Power BI (live dashboard) connection and last but not the least, Twitter connection.

Here, as I am using an Office 365 common account, so my power BI already authenticated using that.

Indiandotnet_Twitter_3

Step 4

Now, let me connect with My twitter account first. Which can be done by clicking Sign In. In this, you need to authorize the flow to access your twitter account.

Indiandotnet_Twitter_6

Step 5

In the next step, you have to define text analysis connection for which you have to first login into Azure and search for cognitive service. If it does not exist, then create by clicking Add button.

Indiandotnet_Cognitive_Service

When you click add button, you will find the below link as you can see. Search Text Analytics and select the option highlighted as below:

Indiandotnet_Cognitive_Service_2

Provide proper Resource Name and select pricing tier as per your convenience. I am choosing free tier for this example.

Indiandot_Net_Vision_API_1

Step 6

Once you create the account, you need to copy the Key from Get your Keys option and copy the Web API option from 2b as shown below:

Indiandotnet_Cognitive_Service_3

Step 7

So far so good, we have created cognitive service and now moving back to Microsoft Flow and configure the Text analytics. Give any name in connection name and provide Account key which we copied earlier and also copy the Web API.

Indiandotnet_Twitter_4

Step 8

Now, we are done with the connection. The next step is to configure the steps. So, when a new tweet is posted, we will provide that tweet to our Cognitive service using detect sentiment which will provide sentiment score. The sentiment score will be given to Power BI.

Indiandotnet_Twitter_8

Step 9

Now, click on twitter option here, you can add multiple text option. For example, you can see in the below search text that we have added NarendraModi, Chokidar or MaybheeChokidar.

Indiandotnet_Twitter_9

Step 11

Now, delete Detect Sentiment action and add again by searching the sentiment:

Indiandotnet_Twitter_10

Step 12

Here, we will configure the sentiment option as you can see. We are giving Tweet Text and Language (we are selecting English Language).

Indiandotnet_Twitter_13

Indiandotnet_Twitter_11

Step 13

Now, in the next step, we will configure the Power BI for which we will login http://app.powerBi.com and click on Streaming Dataset option because we are going to provide.

Indiandotnet_Twitter_12

Step 14

Now, click on API as shown in the below image:

Indiandotnet_Twitter_14

Step 15

Now, define the dataset as shown. Below here, we are defining the fields. So, in the fields, we are adding SentimentScore (which will be given by cognitive service), TwitterDate (when tweet posted), username (who posted the tweet), tweetText (what is the tweet).

Indiandotnet_Twitter_15

Once you define the dataset, click on Create.

Step 16

Now, save the Report and configure the report as per your requirements.

Indiandotnet_Twitter_16

Step 17

Here now configure, Power BI dashboard and remove the payload as shown in the below figure:

Indiandotnet_Twitter_17

Step 18

See the below configuration which configured:

Indiandotnet_Twitter_18

Here if you see, we defined workplace which we have created and dataset defined just above and mapped the fields.

Indiandotnet_Twitter_21

Step 19

Now, save the flow and configuration and try to run it. If everything works perfectly, you will get all the checkboxes.

Indiandotnet_Twitter_20

Step 20

Now, you can see the Power BI dashboard. You will get the graph updating per minute with the new tweet text whenever there is a post with a specific tag.

Below, you can find sentiment score graph.

Indiandotnet_Twitter_22

In the next post, I will share the details of this sentiment analysis data.

Hope you might like this easy way of sentiment analysis.

Happy learning!

License

This article, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Share

About the Author

Rajat-Indiandotnet
Technical Lead
India India
No Biography provided

You may also be interested in...

Pro
Pro

Comments and Discussions

 
-- There are no messages in this forum --
Permalink | Advertise | Privacy | Cookies | Terms of Use | Mobile
Web03 | 2.8.190424.1 | Last Updated 23 Mar 2019
Article Copyright 2019 by Rajat-Indiandotnet
Everything else Copyright © CodeProject, 1999-2019
Layout: fixed | fluid