Using Azure cognitive services to check sentiment in tweets about you can help you find any negative news about yourself
Cognitive Services
Today I had a play with the cognitive services available in Microsoft Azure. Microsoft Flow has an easy to use connector that talks directly to the cognitive services.
Therefore I decided to have a look at sending myself an email as soon as someone tweets in a negative way about my SharePains blog. Am I asking for trouble here?
I’m going to start my flow by collecting all tweets that include the word SharePains. Often people tweet about my posts without mentioning my handle @PieterVeenstra and therefore I’m missing these tweets.
Using the tweet texts I can now identify which tweets are positive and which ones are negative. All I need to do is send the tweet text to the Cognitive services in azure.
Rather than searching for SharePains, you could even look for sharepains.com, which will find all tweets that link to my site.
Sentiment checks
The Detect Sentiment action will now return a number between 0 and 1. Anything above 0.5 is positive. Anything lower is negative.
I single check with a condition will now tell me if a tweet is positive or negative.
Ok, this is great to identify trouble but you could use this in other ways as well.
Do you send out tweets from your company account? Can anybody potentially write these tweets? Do you go through an approval process?
We all want positive news to be sent out on social media. Using the cognitive services you could now check your tweets before they get approved if they are considered positive or negative. This could give your tweet editors the opportunity to validate their tweet before it gets approved.