Startup SemantiStar sees a place in the world of InfoApps for semantic technologies. InfoApps, as SemantiStar CEO Bill McDaniel explains them, present information to the user as needed, when, and in the manner in which they want it. They’re generally are associated with the world of mobile devices – think of many of the goodies you download from sources like Apple’s App Store or Amazon’s Appstore (trademark disputes aside) or the Android Market.
Add semantic technologies to the mix, and these apps can become more context-aware — recall the Siri virtual assistant, which was acquired by Apple last spring. More opportunities exist around the idea of mobile information presentation, leveraging the devices’ geographical coordinates, the user’s historical behavior, and other knowledge that might be pertinent to an initial information quest, McDaniel thinks.
“Traditional applications are not portable – an application on a notebook doesn’t know it’s moving around, but the iPad and Android [devices] have that almost intrinsically,” he says. “There’s information for orientation and acceleration, so augmented reality features can be brought in to add more context and display to what the user gets.”
Clearly combining mobile and semantic technologies is an area where vendors see opportunities, including think of News360, which uses the iPhone’s geo-targeting features to localize news and semantics to categorize and classify it into related clusters. Now SemantiStar is about to release its first mobile device InfoApp, which it’s building for another startup, StreamGlider. It will be a news reader that, among its capabilities, will let users automatically get breaking news related to their current location. “It has the ability to say I know where you are and here’s what’s happening there,” McDaniel says.
Users also can choose to get news around a different location and build their own heterogeneous information sources that play across the screen in real time. Additionally, what users have looked at before will semantically feed into the app’s idea of what else they’ll be interested in, so it can start to automatically assemble relevant news for them. The app will repeat the stories every 30 minutes unless something breaks, related either to the streams they’ve requested or to their current locale or to what it gleans of their interests, at which point it will deliver a new edition. “We produce this so this information floats to the top of users’ awareness. It may repeat over time but the user decides how long the streams are, and how focused or defocused they are,” he says.
“To make it fly, semantics is important. For me to describe and the system to learn by watching what I look at and don’t look at means a better focus on what things are important to me,” McDaniel says. That’s especially the case given that most people don’t have the time to tune and shape the information delivery experience. “This gives you the ability to build channels, as many and as long as you want, and adding semantics lets the application build channels for you and make sure they are more appropriate and more tuned for you,” he says.
The possibilities for semantically-driven InfoApps like this one extend to having those interest models also draw from comments or annotations users make about stories or the streams they share. The company also is working with specialists in the machine learning space to deal with the issue of getting news about something that seems related to a user’s interest but isn’t. Real world experience: McDaniel wanted to get news related to holidays but somehow news related to a port star named Valentine slipped in. “Semantic technology can help weed those things out,” he says.
McDaniel will talk about the direction of InfoApps at the Semantic Technology Conference in June.