Your social listening tool uses plug-and-play AI from Microsoft Azure to measure the sentiment of Tweets mentioning your brand, and alerts you when the sentiment is negative. However, you may wish to include custom ‘negative words’, for example if you are listening for Tweets in a language not supported by Azure’s Text Analytics.

To add custom ‘negative words’ to your solution:

1. Open your SocialListening-checkText Logic App from the Azure Portal.
2. Open the Logic App Designer.
3. Hover between the actions Condition and Flag negative or inappropriate content and click the + icon that appears. Select Add an action.
4. Search for an select the Condition action from the Choose an action box.
5. In the condition box you can add as many ‘negative words’ as you wish. For each word you want to filter for, click on Choose a value and scroll down the the list of Dynamic Content that appears until you find Text under When a HTTP request is received.
6. Set up the condition: Text contains [your chosen word].
7. Change the And/Or dropdown to Or so that the condition will be triggered if any of your list of words are detected. Click + Add > + Add row to add another word.
8. Once you have built your list, within the If true box click Add an action.
9. Search for and select the Set variable action.
10. From the Name dropdown, select negativeSentiment, and for Value enter true (this must be all lower case with no spaces either side).
11. Save the logic app and exit.

That’s it - you’re all set up!

What does the extra action do?

You have added a condition that checks the text of the social media post to see if it contains any of the words on your list. If one or more of the words is detected, the post is flagged as negativeSentiment and you will be alerted!

Read more about Logic Apps condition functions here.