[Featured API Series] Chatterbox, Sentiment Analysis for Social Media
We had the opportunity to ask Chatterbox co-founder, Dr. Stuart Battersby (right), on how his startup’s Sentiment Analysis service is different from the rest. He also explains how Mashape enables him to focus on rolling out new features, while monetizing the API.
What is Chatterbox?
Chatterbox is a Wayra London company, and a spin-out from Queen Mary University. Using proprietary algorithms, Chatterbox identifies and understands brand (b)advocates in social media. Whilst Chatterbox can be accessed via a web browser, integration is very important to our customers and we feel that a flexible service gives Chatterbox a new USP. The earliest component developed at Chatterbox was multi-lingual sentiment analysis designed specifically for social media. Unlike traditional sentiment analysis services, Chatterbox can deal with the slang & informal language, including emoticons and hashtags, that is ubiquitous in social media.
With a critical development roadmap focused on building the best algorithms and IP possible we did not want to have to build user management, billing and rate limiting necessary for a production API. Mashape was appealing as it handles all of this whilst remaining standards compliant. This enables the team to focus on developing and delivering our core service. Integrating with Mashape was trivial, our API already delivered JSON, we simply configured our API with some text descriptions and we were done! Integration took hours to days, rather than the months that a bespoke system would.
All this has enabled us to successfully monetize our API – classifying over 10 million texts in the last 30 days alone – and, as we’ve been able to focus on new development, we will soon be updating our API with Chinese sentiment analysis for social media, and anger detection for social media.
If you haven’t yet, make sure to check out the Sentiment Analysis for Social Media API and give it a spin