The U.S. Patent and Trademark Office (USPTO) has granted more than 9 million patents and, until recently, trying to sort through those and find any patterns or insights would have involved looking through massive amounts of seemingly indecipherable spreadsheets and files.
However, thanks to an open data initiative, the USPTO is allowing users to create visualizations of its data, sorted by geography, date, type and other factors.
In an interview with FedTech, Thomas Beach, senior advisor and acting portfolio manager of digital services and data analytics program at USPTO, says that while the agency’s mission is to deliver timely, quality patents and trademarks, it wanted to explore how to make the best use of its huge treasure trove of data. “And it’s a bit of a leap to go, well, there’s a pile of it, there’s a user community, how do we build the bridge between the two?” he says.
USPTO wanted to open up its data in a meaningful way and also spur future innovations, Beach says. The project the agency embarked on last year not only packages patent and trademark data in unique ways but also was designed to increase understanding on intellectual property and make that data fun, interactive and shareable, he says.
A New Way of Looking at Patent Data
As part of its efforts to fulfill President Barack Obama’s 2012 directive to federal agencies to deliver better digital services, the USPTO’s Innovation lab started a project last year to create a Developer Hub that would make patent and trademark data more accessible.
“We really want to lower the barrier or actually buoy up the population to understand what patents and trademarks mean,” Beach says. “One great way to do that is through visualizations.”
A handful of USPTO employees in the Innovation Lab worked over the past eight months to create the Developer Hub, which officially launched in late April.
Using Tableau Public visualization software, open-source Drupal 7 software modules and other open-source tools, the team created the ability to take patent data dating back to 1790 and trademark data dating back to 1870 and manipulate it in numerous ways.
All of the work was done in-house at USPTO, Beach says. The agency has focused on using agile development processes and DevOps to speed up the delivery of digital services. Doing the work in-house also saved the agency money.
“It takes a great team, I have to tell you,” Beach says. “They’re a very small, scrappy team. My job, I think, is to clear the path and inspire them. Because they’re the experts at what they do. That’s one of the things I’m most proud of, which is creating the ecosystem where something like this can be done.”
Creating Data Visualizations
Much of the USPTO’s patent data had been stored on its website in what are known as “flat files,” which come in the form of HTML pages or PDF reports. The files, created year after year, contained data on all of the fields of categorization that the agency tracks, including filing location, patent type and technology.
Beach says the USPTO “realized very early on” that it needed to create a data hub that would be a destination where users could access the information they wanted. The hub also includes a space where users can share their visualizations and comment on them.
To “prime the pump” and get ideas for visualizations flowing, the USPTO created about 40 different visualizations using the Tableau software, some of which are available now and others which will be released in the coming months, according to Beach. One of the visualizations charts Apple’s financial data with its utility patent filings, and details when the company’s key products were introduced. Another shows the most popular technology categories for patents for each state. Yet another shows patents filed by different universities, allowing friends and rivals to stake claims on which of their alma maters have been more inventive over time.
“We wanted to build a hub to provide you the opportunity to discover your own data story,” Beach says.
The new tools also fit in with the USTPO’s broader mandate, and will hopefully inspire future inventors, according to Beach.
“Data discovery on intellectual property is sort of like the core mission,” he says. “It doesn’t matter who you are, what you’re like, what your education is, where you come from and where you live in the world. If you have an interesting idea, you should be able to have a low enough barrier to entry to actually figure out, ‘Is this something that is worthy of getting a patent or a trademark on?’”