EU-funded project to prompt intelligence-sharing

While the information is in the open - either in traditional sources such as newspapers or radios or online in blogs and social networking sites - harvesting and analysing so-called "open-source intelligence" in order to better anticipate or manage events is hampered by a lack of compatible technology.

Virtuoso (standing for the unwieldy Versatile Information Toolkit for End Users Oriented Open Source Exploitation), when ready, is to be a one-stop platform for the use of all analytic tools such as speech-to-text software, tools that show where people are speaking from or tools that process proper nouns and locations.

"What we are developing is the integration framework where you can plug in all the different tools that you might use and they can all communicate with one another," says Axel Dyevre, from European Strategic Intelligence Company (CEIS), one of the partners in the project.

He compares it to Microsoft's Windows system. "You have Word and you have Excel and if you take an Excel spreadsheet, you can easily import it to Word."

With a budget of €11.45 million, of which €8 million is EU R&D money, the project has 17 partners including TNO, a Dutch research organisation, Denmark's Aalborg University, the Austrian company Sail Labs (specialised in speech to text technology) and Newstin, a data acquisition company.

The first big test will be in June when three scenarios - border security in Greece; arms proliferation and a potential attack on member-state infrastructures - will be used to showcase the project.

Watching will be a roomful of officials from the EU border agency Frontex, the EU crisis centre (SitCen), the EU satellite centre as well as experts from national intelligence agencies, military and police personnel and foreign and defence ministry officials.

Secrecy and prejudice

But the project has to fight against both the culture of secrecy among intelligence agencies and a certain prejudice towards information gathered from public sources.

"There can be some paranoia among those working in intelligence. For them, keeping the information is much more important than collaborating even it if would be useful [to share it]," says Frederik Schumann, a consultant with CEIS.

"For them, open source [information] is something that kids play with," he adds.

But Virtuoso could help with analysis of situations in hotspots the EU is involved in, he argues. He points to US military personnel in Afghanistan who have said that local newspapers provide more valuable information than their own intelligence services.

The current war in Libya is a case in point. Experts remain unsure about the extent of the bloodshed, where loyalties lie, how the different tribes interact and the effectiveness of the international coalition against Colonel Gadaffi's forces.

"Thousands of people are writing about [Libya]. There is TV, radio, bloggers. There is so much information. The idea is to extract what those sources don't talk about," says Schumann.

Dyevre, formerly an officer in the French army, is hoping intelligence services will ultimately use Virtuoso to swap information directly with one another: "Another interest is, but later I think, the possibility of information sharing between organisations if they are using the same standards. [Exporting] a whole set of data and [giving] them to another organisation."

Whether their project will eventually overcome intelligence agencies' ingrained reluctance to share information is moot.

But both Dyevre and Schumann are adamant that their project will not go the way of many schemes funded by EU money - to a dust-gathering shelf full of worthy but impracticable research.

"The project is quite novel in that it is very user-centric and we are working very closely with the end-users," says Dyevre. "We want to be a FP7 (framework research programme 7) project that actually produces something," adds Schumann.

Meanwhile, another similar venture is already in the pipeline. Member states have just said yes to EU funding for a project aimed at reducing cognitive biases – analysing data according to the data gatherer's world view - in intelligence gathering.