Commerce Data Chief: Tech Can Make the Patent Process 'A Hell of a Lot Easier'

The Commerce Department's chief data officer Ian Kalin

The Commerce Department's chief data officer Ian Kalin Flickr user Kevin Krejci

Patent applications could be processed faster if patents were machine readable.

Systems that can interpret natural language could speed up the patent application process, but Commerce Department's old-fashioned routines stand in the way, Chief Data Officer Ian Kalin said Wednesday. 

Kalin, who took office in March 2015, told an audience in Washington the "entire patent system is in unfortunate contradiction with the general principles of modern data science." 

"The basic obligation of a patent examiner is to prove that something does not [already] exist," he said at an FCW event, explaining that existing patents are published online as picture files whose text is nonsearchable.

 “We have machine readable versions, we just don’t publish it," he added. "And that's not even big data; that's just basic customer service."

Using more advanced search mechanisms such as natural language processing could be used "not to obviate or remove all patent lawyers, but to at least make their jobs a hell of a lot easier."

Kalin said the U.S. Patent and Trademark Office has an extensive "data road map" that includes plans for publishing, and making use of, data collected by the agency.

But "there's a history, there's a process" and sometimes, cultural barriers that prevent the agency from acting on that road map, he added. 

The patent application process could be improving, however slowly. USPTO last month unveiled Dossier Access, an online system designed to help patent seekers search the dossiers for related applications filed at other offices.