Sentiment Analysis Using Knowi and AYLIEN

I had a phone call from a customer the other day. Went like this:

Steve: “I know you do Machine Learning n’ all, but do you do sentiment analysis? We track the feedback from our customers inside the venue and we want to track the experience for each of our events and combine with our data from our database”

Me: “No, we don’t do Sentiment Analysis out of the box. But, if there’s a service with a REST API that does sentiment analysis, then we should be all good.”.

Did a quick google search and enter AYLIEN. They offer a SaaS-based application that can be called via an API for intelligence on textual content, including sentiment analysis.

With Knowi, we can combine REST API’s with database data to join the data together that can be used to visualize and analyze along with sending it to downstream systems.

The next steps were straightforward in Knowi:

Query the natural language feedback data from the database.
Join it with AYLIEN API to determine sentiment for the feedback in natural language.

The Aylien API requires a separate call for each text to parse, so we’ll use a Loop Join join type that allows for..each type of type of Join within Knowi, with a parameter. The {text} above in the screenshot is the parameter.

Results:

That was it. Within 10 mins we had a fully functioning sentiment analysis capabilities leveraging AYLIEN and Knowi’s multi-source joins.

Jay Gopalakrishnan is the founder of Knowi, an AI Driven Analytics platform for Modern data architectures. Follow Knowi here.