Yes, we are boiling the ocean on all topics to provide both a microscopic analysis of sentiment and its drivers as well as a 50,000’ view of the landscape to see what sentiment is bubbling up.
Our ubiquitous topic approach increases accuracy by focusing sentiment on how people speak as opposed to individual words. In our world the tweet “I love the LA Dodgers” is only a positive for the baseball team and not for the city of Los Angeles or those avoiding taxes.
We also serve up related topics of discussion and sentiment for each so you’re immediately able to see if sentiment on Trump is up based on a recent rally and down based on Chinese tariffs. Or compare opinion on smartphones across their battery, camera, OS, screen, and applications. These related topics can be directed to look for sentiment on specific issues and undirected to see sentiment on issues bubbling up of which you may not be aware.
Our approach allows also us to remove irrelevant related topics and sentiments from results and immediately see revised results with no new processing.
So, removing “pie” from results on Apple the company, tossing “suspended” as a sentiment from results on Twitter’s suspension of Mitch McConnell’s Twitter account, or even reverse the polarity of a sentiment when the context calls for it, can be done with a click of a mouse.
Leaving Booleans behind allows us to deliver unprecedented speed, accuracy and flexibility unavailable elsewhere.