Summary:

Today our smartphones know where we are, what time of day it is and, in some cases, our likes and dislikes. Seymour, a new self-learning recommendation engine, leverages that contextual information and combines it with the collective intelligence of the Internet for real-time, useful recommendations.

seymour-featured

Today our smartphones know where we are, what time of day it is and, in some cases, our likes and dislikes. So why can’t those contextual data points be effectively combined with the collective intelligence of the Internet in order to provide real-time, useful recommendations? Part of the issue is sifting through the massive amount of available information and bringing it to a personal level. Seymour, a personal concierge app for Microsoft Windows Phone 7 devices, is taking steps to make this vision a reality by empowering users with a smart recommendation engine.

Seymour sounds much like Siri, the personal assistant for iPhone founded in 2007 and later purchased by Apple. I expect more similar competitors to arrive as the challenge of mining through a growing amount of data will only frustrate consumers. And we’re already seeing recommendations become more integrated with location-based services, but that’s only part of Seymour’s intelligence. Seymour — the brain child of Clever Sense, a start-up with $1.6 million in funding — includes both the end-user application and two back-end technologies: the Extraction Engine and the Serendipity Engine.

The first solution uses natural language to help translate unstructured information on the Internet into understandable human terms. This helps Seymour’s ability to understand natural queries such as “What’s the best Italian restaurant nearby that has an open table at 7 p.m.?” instead of requiring users to search by keywords that may or may not provide relevant results. The Serendipity Engine pulls in the contextual data already available from a smartphone or user habits: location, time, intent and more. So firing up Seymour in New York for a food recommendation at 8 a.m., for example, will result in breakfast options in Manhattan instead of suggestions for dinner in San Francisco.

This video demonstration of Seymour impressed me, and I’ll soon be loading it on the HTC HD7 Windows Phone 7 device I won on eBay last week, due to arrive this morning. The application not only learns user behavior on its own, but can be tweaked through a familiar “like” or “dislike” button. If you don’t care for Seymour’s results, you can simply tap the thumbs down and the software intelligently modifies its personal recommendation engine. The suggestions actually appear to get better through crowdsourced information as well because Seymour scours web-based reviews of places, products and services to offer the pros and cons of its recommendations. Babak Pahlavan, the president and CEO of Clever Sense, explains the technology Seymour uses here:

Although most consumers will see benefit from recommendation engines such as Seymour, Siri and the like from the obvious use case, I can’t help but think forward to my vision of home robotics as the next big thing. The guts of a smartphone embedded in a simple robot could act as a truly mobile and personal concierge at home when fitted with these self-learning, artificial intelligence web-connected platforms. I’ll probably have to wait a long time to see that happen, but in the meantime, users of iPhone and Android devices won’t: Seymour is slated for those platforms in the near future.

Related content from GigaOM Pro (subscription req’d):

You're subscribed! If you like, you can update your settings

Related stories

Comments have been disabled for this post