Canvs’ patented model of 42 emotions goes beyond basic sentiment to unlock a more nuanced understanding of consumer feedback
Canvs AI and Fuel Cycle announced that Fuel Cycle has added Canvs MRX to the Fuel Cycle Exchange (FCX) Marketplace for text analytics. Through the partnership, Fuel Cycle customers gain access to Canvs’ innovative “bring your own data” solution for analysing unstructured text data from virtually any source, including survey open ends, customer reviews and social media.
“Companies have virtually unlimited amounts of open-ended text responses from customers available to them but face a significant challenge in generating meaningful insights because of the time, complexity and cost of analysing it all. Being selected for FCX makes it easier for Fuel Cycle customers to embrace the power of Canvs to analyse open-ended text with a high degree of accuracy and control. The result isn’t just faster time-to-insights; it’s more comprehensive, better-quality insights,” said Jared Feldman, founder & CEO of Canvs AI.
Canvs dramatically accelerates the time required to code responses to open-ended survey questions. Advanced, AI-powered Natural Language Processing (NLP) delivers remarkably accurate classification of topics and emotions expressed in unstructured text, along with a customisable code framework. Powered by AI and machine learning that was trained on social media, providing a rich ontology based on the way humans actually communicate (emojis and misspellings included), Canvs’ patented model of 42 emotions goes beyond basic sentiment to unlock a more nuanced understanding of consumer feedback.
The platform utilises a robust drill-down architecture allowing researchers to quickly and easily dig into the insights from open-ended text. It can analyse both open- and close-ended responses to identify statistically–significant findings in data automatically. Additionally, powerful Boolean search capabilities allow researchers to quickly search and filter based on emotions, topics, codes, text or closed-ended questions.
