The question: how does NASA automate the process of identifying concepts, people and other entities in NASA’s intranet content? Not an easy solution, given the vast amount of information this agency houses. Posted on the NASA@work crowdsourced platform—a place for collaborative and innovative conversations—this quandary was spotted by two Innovim software engineers—Larry Gilliam and Ali Rezaiyan.
“While the guts of this solution is a bit Greek, there were two challenges for us: the first to index and manage contents—essentially cataloging the information, and the second (which is the real challenge) was to understand user intention with they provide a general search term,” says Gilliam.
The example Gilliam and Rezaiyan used in their solution was the term “eagle.” The music band? A sports team? Or the animal?
“Query disambiguation—or simply figuring out an ambiguous search term—is tough business. It would be like if you went into the library and asked the reference desk for that book about the sharks and the scientist. That’s a bit of a hyperbole, but how does NASA provide the best answers to their users searching for, at times, vague agendas? That’s what Larry and I were trying to tackle,” says Rezaiyan.
Their plan and architecture for the indexing and querying process was the winning formula. As such, they will be receiving a Certificate of Excellence signed by the Chief Technologist, and are eligible to win one of the NASA@work awards found here: http://goo.gl/ehbsvo.
We congratulate Larry and Ali for their excellent, creative, and disambiguating work!Share