Army intelligence to stop roadside bombs came from an unexpected source: an experimental system that organizes unstructured data on officers' social network.
Neal, an enlisted man in the U.S. Army on a tour of duty in Iraq, came up with a new way to detect improvised explosive devices, the notorious IEDs that plague convoys and troops on patrol. Neal's expertise was largely unknown until it was discovered by an experimental system applied to a social network used by Army officers.
Thursday, one of that system's authors, David Gutelius, now chief social scientist at Jive Software, told attendees at the GigaOm Net:Work conference in San Francisco that "machine learning" could be applied to social networks used in the enterprise and yield high value information there as well.
"Enterprise problems are really quite similar to some of those encountered in the military," where, in the face of a challenge, new expertise needs to be discovered and widely deployed fast, he said at the event.
The IEDs were causing a fresh round of casualties deep into the eight-year Iraq war, after militants devised a new way to conceal and set them off. Captains of Army units were repeatedly discussing the problem, but no one quite realized that an individual in their midst had found an answer.
The Defense Advanced Research Projects Agency, however, had funded a project to produce a Computer Assistant That Learns and Organizes unstructured information. The tool, dubbed CALO, was useful to apply to a social network such as Online Command, a community the Army had created for its captains where they could talk online as peers. In the midst of an active war, the Online Command social network produced a lot of ill-defined and unstructured data. IEDs themselves are loosely defined as amalgamations of old artillery shells, mortar rounds, or other explosives to form a roadside bomb with a deadly amount of force.
The CALO tool was applied to the Army captains' social network and built a reference list of Neal's connections to successfully foiled IED attacks. It noted the importance that participants attached to information about Neal and the frequency with which an officer forwarded an experience that included Neal's name. With this information, CALO moved from passive onlooker to active commenter itself. When questions about IEDs came up, CALO advanced Neal's name.
Gutelius, an artificial intelligence expert then working for SRI as part of the CALO project, thought the system had gone bonkers when it kept advancing Neal's name on the subject of IEDs. Gutelius was watching how it performed; he asked himself why such a sophisticated system was coming up with the same simple answer over and over again.
"It kept noticing this one guy. It was just hammering on Neal. I thought it must be broken," he recalled in an interview after his GigaOm talk.
He pointed out the "defect" to a group of Army captains at West Point who had just returned from a deployment to Fallujah. One of them said, "I know Neal. He's awesome."
The Army captains helped him understand the significance of the information that Neal (whose full name can't be disclosed) was associated with, and then shared that intelligence with the Army's high command in Iraq. "Suddenly, it made sense what the system was doing," recalls Gutelius. Neal's expertise was captured and spread throughout U.S. combat units. Neal, meantime, was whisked out of Iraq to Fort Leavenworth in Kansas, where he became a founder of the Army's Center of Excellence on countering IEDs.
Gutelius said the high point of his CALO experience came when he went to Fort Leavenworth to do a presentation on the use of intelligence from social networks. He cited the IED gains made from analysis of Online Command data. After his presentation, a three-star general told him he knew the same system was saving lives in Iraq. Also, Neal introduced himself to Gutelius; he hadn't realized to what extent social networking had affected his life until hearing Gutelius' presentation.
"Enterprises are not bounded entities," said Gutelius in the interview with InformationWeek. "They are emergent organisms," particularly now, as previous, strictly hierarchical ways of doing things are giving way to flatter, more dynamic teams and project organizations. As conditions change around a company, an analysis system applied to an active social network "may be able to extract a different kind of value out of the human interactions going on" than traditional management techniques, he said.
Jive, SAS, and other companies are building such analysis systems as a way to pursue social network intelligence. Another set of companies is attempting to define standards for social networks so that analysis tools might be applied across them.
Is your company antisocial? Our latest research shows that business-oriented social networking platforms aren’t living up to their promises of better communication, collaboration and productivity. Download the report here (registration required).
The Agile ArchiveWhen it comes to managing data, donít look at backup and archiving systems as burdens and cost centers. A well-designed archive can enhance data protection and restores, ease search and e-discovery efforts, and save money by intelligently moving data from expensive primary storage systems.
2014 Analytics, BI, and Information Management SurveyITís tried for years to simplify data analytics and business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?