Senseforth: extracting wealth of crowd wisdom trapped in online unstructured content
In this age of the internet, there is a huge amount of unstructured content online like social media posts, web pages, blogs etc. Search technologies do a good job in helping “find” information, but not really when it comes to making sense of it for the purpose of decision making, believes the team behind Senseforth. Founded by Krishna Kadiri (CEO), Somakumar Kolathur, Sreekumar P, Sridhar Marri (Advisors) & Ritesh Radhakrishnan (CTO), Senseforth aims to help people make better crowd wisdom-driven decisions by delivering insights derived from large volumes of unstructured data like text, audio and video.
The founders are industry veterans and ex-Infoscions with decades of experience in the space of internet, information management and user experience. Senseforth.com was launched as public beta in November 2012 but the work on the technology platform has been going on since 2010. “As online content shifts from limited set of organization websites into numerous pieces of user generated content like tweets, blogs, video posts etc., it becomes impossible to read through search results. And hence something like Senseforth makes sense,” says Ritesh.
Senseforth’s algorithms use techniques like Natural Language Processing, Machine Learning, Semantic Search etc. to machine read millions of pieces of text, extract relevant information and generate insights, that is make sense for various decision contexts like shopping, product management, elections, reputation management, CRM and so on. The company has been in public beta for a year and supports purchase decision making for mobile phones and tablets. “Senseforth Shopping Assistant is also integrated into a leading mobile device retailer and is helping them improve user engagement and conversions,” says Ritesh.
One of the toughest problems for Senseforth probably would be to recognize all the genuine opinions on the web. “We have spam filters which remove duplicates, ads/offers and suspicious content. We aggregate content from many sources and across many pieces of content and factor in other things like popularity of the post to reduce the impact of spam,” says Ritesh. The live counter on the website suggests that the Senseforth algorithm has crawled over 60 million opinions from which it makes sense. The current focus of the company is to strengthen the technology. Once perfected, the revenue stream can be developed which will predominantly be to give insights to brands and manufacturers about their products. Reviews42 is another venture-backed company which helps make sense out of reviews.