Jennifer Zaino
SemanticWeb.com Contributor

Some three months into its live public beta, semantic matchmaking site Bintro added the Freebase and Geonames ontologies to the data sets that help it identify and disambiguate information to create highly relevant matches between job-seekers and employers; charities and volunteers; potential partners and joint venture associates; investors and opportunities; and mentors and those looking for guidance.

It's using Geonames to expand its geolocation services. That, says Bintro CEO Richard Stanton, will make it easier to match, for example, someone who express the desire in its web form's unstructured text fields to work in one of New York City's five boroughs and an employer who posts a position located in the Bronx matching that person's skills set.

Stanton isn't specifically disclosing how Bintro plans to use Freebase ontologies other than to say it's helpful in building strong lists that are specific to its forms. But he notes that Bintro is also interested in other data sets in the way of broad-based ontologies, such as the OpenCyc content, as well as some individual specific taxonomies or even glossaries that it may leverage to provide its services

You don't even have to be there

Since its launch May 13, the service has generated thousands of matches, Stanton says. What's drawing users in is the seamlessness of the experience, thanks to its combination of semantic and proprietary technology. Once a person uses what appears to be a standard web form to input their information, such as the kind of work they are seeking, using both menus and natural language commentary, there's no need to update it or search different lists for jobs; neither does the potential employer have to cull through hundreds or thousands of biographies to figure out who to talk to.

"We do that work in an automated fashion," Stanton says. "Our matchmaking occurs without your even being on the site."

Matches are made based on the merits of who the posters are and what they have done and can offer, not the date and time they entered the information. So, a job seeker can create a posting and even if there are no current employers looking for those skills, as soon as such an employer lists an opening the match to that job seeker is made.

"We're leveraging the semantic web for bringing people together to express and fulfill needs," he says. "You see classified data on [other] sites and they are antiquated. It's a long list and nothing about the site is continuous or automated in relation to the user."

The narrative fields are where the semantics come in with Bintro, Stanton says. Here, a job poster who checks off "accounting" from the menu options asking about her area might describe herself as Black-Scholes expertise; someone searching for assistance in evaluating opinions on options might never use the term "Black-Scholes" in his own narrative field, yet Bintro is smart enough to make the connection to bring the two parties together because it can disambiguate terms based both on publicly available and proprietary ontology resources.

"We know the category you are in but we are interested in mining the specifics of what you are telling us to give you a highly relevant match," he says. "That's how we leverage the semantic web."
The company's plans are to continue to learn the nomenclatures behind different categories to continue to drive relevance in results. That said, users don't know they're dipping into the semantic web when they take advantage of the service-nor do they have to. "We're not dragging the user into the semantic aspect of it," he says. "They just use a web application that looks like what they were using before."

Bintro, which should emerge from its live public beta state at the end of the fourth quarter, is one of the first large-scale consumer applications based on the semantic web, aside from services such as search or advertising models within publishing platforms, he says. There is potential for its matchmaking capabilities to benefit other sites, such as some smaller social networking sites with an interest in driving revenue through new premium services, he acknowledges. But for now Bintro is most invested in enabling its own consumer portal - which in the late fourth quarter probably also will add fee-based premium services to its lineup; right now ads support the free basic consumer services.

Bintro also has built an enterprise-oriented version that mid-sized and large companies can use internally, using proprietary forms specific to their needs to enable matching between corporate and business line needs and the employees who have the skill sets to fulfill them. In bigger companies, especially those with worldwide presences, making those matches is harder than you might think. Today they try to make those matches with email queries - but Stanton says there's a better way.

"The essence of what we do ism we are a matching engine. We are not about building social networks for organization and we're not about simply managing knowledge. We're about producing tangible matches. ...We say we want to give them the end result, an actual match."

The company also plans to add more matching categories over time, adding to its people-focused categories matching services around places (real estates, sublets, room-mate finding, etc.) and products.