Editor’s Note:This SFNewTech demo-highlights post is from our July 20th 2010 Event. Please bear with us while we play “catch up” on our shiny new blog 🙂
Search engines, social networks and online discussion forums represent much of what’s good about the internet, while trolls and online predators who stalk and prey on innocent victims with tremendous ease represent the potential bad.
A quick online search shows that online predators are steadily on the rise. So much so, that Dateline NBC’s Chris Hansen devoted an entire show to baiting and arresting online predators in his shockingly disturbing, yet wildly popular “To Catch A Predator†T.V show.
So how can website owners, online social networks and discussion forums control this problem? Crisp – founded in 2005 by Adam Hildreth – has built a sophisticated and effective system just for that.
Crisp CTO Peter Maude came on stage at our July 20th SFNewTech event and presented their service. Using their proprietary behavior pattern tracking technology, Crisp collects pattern information that makes a site safe, thereby protecting a company’s brand (think Disney, which caters to kids). Crisp detects behaviors, then profiles and auto-manage users to keep them safe online with its unique, purpose-built systems for MMOs (Massive Multiplayer Online games), virtual worlds and social networks. Yes, that’s a mouthful, but that’s what’s needed to catch these ever evolving predators!
Maude showed us how Crisp uses specific word and phrase filters, generated by customer request, to detect patterns of abuse across every user profile on that customer’s site. If Crisp detects an abusive or spamming pattern, their system instantly blocks that abusive user’s typed entries (in say a chat room or discussion forum) thereby saving other users from the nuisance of universal web abuse. In the case of minors, it prevents the ‘grooming’ of children by online predators by analyzing behavior patterns using syntax and language style, thereby protecting children and minors online before the predators can make a connection.
And Crisp does all this in real time, with a response turnaround of about 100 miliseconds! Maude demonstrated this using a simulator on his laptop where he entered a chat room and typed abusive keywords and phrases and showed how the Crisp system intercepted and blocked those messages, and the abusive user, in real time.
Maude noted, “many web sites need to censure based on a multitude of what they consider to be their own threat vectors” and that is exactly what Crisp delivers.