The Department of Energy (DOE) recently released a nifty patent visualization tool – the “Visual Patent Search.” The tool showcases all US patent applications and issued US patents created through DOE funding. Users can analyze patent documents by individual DOE labs or by patent status.
The software behind the Visual Patent Search is based out of two Pacific Northwest National Laboratory (PNNL) innovations: IN-SPIRE™ visualization technology and the Scalable Reasoning System (SRS) text and data analytics platform. This technology is similar to the UC Berkeley Flare software incorporated into this BBC graphic on internet usage trends.
The Visual Patent Search categorizes DOE patents into broad technology groups, from energy and computer sciences to life sciences and infrastructure. The size of each square depends on the number of patents within each category, giving users an intuitive grasp of the broad trends in DOE funding. By clicking on any square, users can access a list of documents within the relevant category (methanol patents below):
DOE’s visualization software, and other patent visualization tools, allow users to quickly and intuitively analyze the size of a patent portfolio, analyze trends in portfolio investment, and see a visual layout of technology hierarchies. The Visual Patent Search is also an iterative process. Given a user-inputted time range or search keyword, the map will rearrange accordingly.
As a result, users can simply click on the technology of interest and isolate patents of interest, rather than wade through thousands of (tedious) patent documents.
DOE’s visualization software provides useful high-level data, but is still too limited for most patent needs. For one thing, the DOE patents only have value in context to similar patents outside of the DOE, and this tool does not allow users to analyze non-DOE and DOE patents side by side. In addition, users cannot see whether a patent falls under multiple technology categories – requiring a read-through of patent claims to see all the potential applications of a patent.
Furthermore, the Visual Patent Search is not intelligent. Patents are categorized by whether certain keywords show up in the patent text, rather than by how they fall intuitively within the categories. For example, under the Energy –> Fuels –> Methanol category, one would expect patents related to production of methanol for fuels. In this category, however, a metal etching patent shows up (US Patent 6,139,716 submicron patterned metal hole etching), as methanol is used as a wetting agent in the process.
Overall, the DOE’s Visual Patent Search is a step in the right direction for patent experts and non-experts alike. By making DOE patent information accessible and transparent to the public, policymakers can see where DOE tax dollars are going, while innovative citizens can license and access these state-of-the art inventions for broad use.