Sentiment Analysis – Using Alteryx + Azure Cognitive Services + Tableau

What better than having complete sentiment analysis done without writing any codes as such for it. Exploring more about Alteryx, next in pipeline was the more advanced features provided by Alteryx for Text Analytics. Having done the same using R and other NLP methods for Sentiment Analysis this was pretty simple and what better having this data ready to be visualized on Tableau.

With Microsoft Azure, Google providing API’s (With very Minimum Pricing, Azure charges $1 for 100K/30 days calls) for NLP it’s just a matter of integrating this within the applications we use for data processing. With Alteryx, connectors to a lot of these API’s are readily available for free at Alteryx marketplace if not available in the trial versions.

Here is a sentiment analysis on the recent goof-up that happened while presenting the Best Movie award at the closing stages of Oscar’s. This analysis is based on the tweets captured related to same. Tweets pull is directly done using twitter search connector available with Alteryx, you just have to provide the API call details and the search that needs to be done. You can also the limit the number of tweets that needs to be pulled in.


This how the flow in Alteryx looks like, with twitter Search connector being the first element to read the tweets through a API call, after further cleansing of tweets and sampling a call to Azure Cognitive Service for text analysis is made which returns Sentiment score for each of the tweets, and results are pushed into a Tableau Data Extract for further visualization and analysis. This whole flow execution just takes few mins to run end-to-end.


From the visualization above it is quite clear that during the last few minutes of Oscar’s people were tweeting positively about the Moonlight-La Land saga.

This is second blog inline with my exploration on Alteryx and it again proved be a self-service analytic tool with so much of features readily available.

Happy Learning !!