In an effort to simplify patent screening Japan to use AI

May 15, 2017

The Japan Patent Office is planning on using artificial intelligence (hereinafter referred to as ‘AI’) in the processing of patents. AI will help to automate ‘cumbersome’ tasks in patent, trademark and design applications, such as literature reviews. The office hopes to begin using this technology to automate 20 tasks by 2018, according to Nikkei Asian Review Report.

In this context, AI will be able to search through lots of files and documents to check if a piece of technology or intellectual property has already been patented. It will also be able to classify patents by field, according to the report.

Furthermore, image recognition technology will be able to verify trade mark applications against previously registered images and logos.

This is not the office’s first foray into AI. Since December last year, it has been using an AI system to answer queries about patents, the report said. The performance of this technology proved that it increased operational efficiency and helped to “curb long working hours”.

Testing of this new AI system will begin this summer, starting with 6 of the 20 tasks, and will continue in stages for the remaining 14 in the next fiscal year.

Artificial Intelligence is the next big thing in the legal profession, especially in an age when corporate clients have become increasingly cost-conscious about their legal bills, refusing to pay for the hours spent on research, even as those hours soar. The bottom-line is, for the legal profession to stay competitive, they must start cutting costs. That means finding ways to make processes like research more efficient. That’s where ROSS Intelligence comes in. Built on the Watson cognitive computing platform, ROSS has developed a legal research tool that will enable law firms to slash the time spent on research, while improving results.

Current legal research offerings like Bloomberg BNA, LexisNexis and Thomson Reuters come with a steep learning curve, requiring training that’s not built in to the billable hour model. In other words, it doesn’t pay to learn how to use these specialized platforms. Internet search is more user friendly, but returns poorer quality results that still need to be sifted through manually.

Not only can ROSS sort through more than a billion text documents each second, it also learns from feedback and gets smarter over time. To put it another way, ROSS and Watson are learning to understand the law, not just translate words and syntax into search results. That means ROSS will only become more valuable to its users over time. Which is not to say that ROSS will be replacing lawyers. Weighing data, drafting documents and making arguments—those will still be left to the humans. But by tackling the burdensome task of research, ROSS frees up lawyers to do what they do best, and helps keep costs down, which brings down the point of entry for clients.

For more information please contact us at :