The Web Innovators Group last night showed a welcome change from the high-tech meetup’s earlier sessions this year. This winter’s WebInno meetings were deluged with job seekers – but last night’s group of about 300 seemed to include more people wearing the event’s trademark “I’m hiring…” name badges, or pitching their own companies in informal networking sessions.
Out of the event’s featured “main dish” presenters, three-year-old Book of Odds Enterprises Inc. was the fan favorite. The company is developing a semantic search site that parses probability statistics from the web and presents them in a format designed for consumers. With a beta invite key, you can find out, for example, that if you live in a city, the odds you average less than six hours’ sleep a night are 1 in 3.5.
The beta site’s search engine seemed to be having trouble this morning: searches for common odds queries – like car accidents, business failures and plane crashes – yield no results.
But the audience at WebInno liked the idea. They voted it tops in a text-message vote, over Batch Book, a Providence-based small business CRM software, and Epernicus, a social network for scientists.
Posted by Galen Moore
Tags: BatchBlue, Book of Odds Enterprises, Epernicus, WebInno
RSS Feeds
Thank you for the write-up and a BIG thanks to the organizers and all in attendance at webinno23. We had a blast presenting our beta site and all the feedback we’ve received is immensely valuable for improving Book of Odds.
One clarification on the post above: Book of Odds is, in fact, not a semantic search engine at all. We do not crawl the web and parse data in real time. Rather, we have a staff of researchers who meticulously vet the data behind the scenes. Once computations have been made the resulting data is then loaded into our ontologies and back-end semantic database (built by Cambridge Semantics) – which is used for organizational efficiencies. It is a subtle point, but an important one: Every data-point that goes into our production process (and there are millions) is evaluated by a human researcher (and much data is thrown out due to methodological issues!), so users can be confident that we are presenting only the best data we could obtain. Keep in mind we’re still in beta and will be fine-tuning our search for as long as we’re in operation – to date, we’ve only loaded about half the content we will have on launch (mid-October) and new content is being created every day, so search results should continually improve.
Thanks again and I hope you’re enjoying the beta!