Web Recipe: Add Pinch of Semantic Flavor
“The Internet is a vast and complex network of websites with huge amounts of different information, from blogs and tweets to news sites, forums, P2P networks and 2.0 applications such as Facebook and YouTube,” says CTO Fernando Zunino. “The present state of things is chaotic and intricate but very promising: It is the dawn of the Semantic Web, the Meaningful Web, where computers will learn to understand human needs and present solutions to every-day problems. We have developed a semantic tool that we hope will take us one step closer to our dream: giving computers the ability to understand, to generate meaning, from human information.” Meaningtool, he explains, can categorize any kind of text, from those aforementioned blog entries and news articles, to those video comments and tweets, while also extracting brands, people and organizations; generating tags for SEO purposes; and, in the near future, returning a measure of the mood or sentiment present in any website. “Developers can use these powerful tools to enrich content and applications with semantic data; we have made it simple to develop using Meaningtool and we have also released a free categorization service to work with,” he says in an email interview. “We want everyone to enjoy the benefits of the Meaningful Web.” At one end of the spectrum—smaller-scale bloggers, for example—the benefit can be Meaningtool’s ability to simply handle the task of categorizing hundreds of entries that have been written over time, be it for SEO purposes or simply for classification and organization needs. “And to make it even simpler, we are working on a new concept using Meaningtool's technology, which will be designed specifically for end-users and small companies,” Zunino says, though he can’t give any further details at this point. Or consider the online publishing and advertising space: Every day, billions of ads are placed on websites without much knowledge as to what those sites talk about, nor which language they are written in or what kinds of people visit them, he says. “Publishers have to process millions of URLs using simple keyword extraction techniques, antique marketing classification strategies or relying on man-power to assume such a complex task,” he says. “The result: Advertisers can't get the product directly to the customer. Meaningtool can process and remember millions of user-profiles helping to deliver the information directly to the customer. No more misplaced banners -- now advertising campaigns will reach the right people in no time. And publishers will get a platform that can categorize and extract meaning from millions of, until now, miscellaneous websites.” Popego is currently working with Ad Networks to help with better ad placement. “In order to do that we have developed an application that integrates with ad servers and tracks their users’ activity,” Zunino says. “Then using that data we apply our technology to build interest profiles of those users and we send these profiles to their platform so they can make an informed decision regarding ad placement, increasing their click-through rate.”
Additional adaptability comes via Meaningtool’s ability to process text in major languages including English, Spanish and Portuguese. Multilingual support, Zunino says, is simply what has to be done. “Diversity will continue to increase in the Internet over the next few years and it is important to develop tools that everybody around the globe can use, no matter if they are in Arizona, Hamburg or Santiago de Chile.”
Email This Post |
The Voice of Semantic Web Business
|
|||||||