June 10-12, 2018, San Francisco, CA: Representatives from IP.com attended Intellectual Asset Management (IAM) magazine’s Intellectual Property Business Congress (IPBC) conference to exhibit our InnovationQ Plus™ software. We partnered with IEEE to best explain how our software has become the industry-leading non-patent literature (NPL) search tool. Conference participants and attendees were industry leaders in the IP sector. We spoke with a number of attendees about our tool and discussed with IP experts how artificial intelligence is affecting both IP protection and discovery. Gillian Micoli, Senior Account Manager with IP. com, shared some of what she learned.
One of the key breakout sessions we attended was The World of Artificial Intelligence. Speakers included William R. LaFontaine (General Manager of IP, IBM), Nicolas Schifano (Senior Director and Assistant General Counsel, Microsoft) and Harrick M. Vin (Founder and Global Head, Digitate). The discussion revolved around how AI is affecting the IP landscape, specifically regarding patent protection. IP.com began investigating this space in the late ’90s. Since then, interest and investment in the development of machine intelligence technology has substantially grown. The panel held interesting discussions, specifically looking at the IP landscape, about the challenges this trend presents and developing solutions to meet those challenges.
To start things off, Vin clarified the differences between machine learning, artificial intelligence, and data mining. All three phrases are now buzzwords—overused by the general public, leaving the overall understanding muddied and the value diluted. As Vin explained, machine learning is a way of learning patterns and adapting this learned knowledge into intelligent systems. Artificial intelligence is the science of training computers to learn and act like humans and improve their learning over time in an autonomous fashion. Data mining is the practice of examining large databases to generate new information. To achieve an optimal result, systems leverage big data and feed the information in the form of observations and real-world interactions.
The next to speak was Bill LaFontaine from IBM. He drilled down into how AI can be leveraged for patent protection and how is this different from keeping trade secrets. Working in software, the need for “explainability” resonated with Micoli. As LaFontaine said, no one appreciates staring into a black box and simply trusting the results. It’s best to walk a client through how the systems react and results occur. When a client sees InnovationQ extrapolate results for the first time, they often have an unsettled response to the “black box.” How can a short, semantic query lead the software to percolate the exact patent for which they searched—and had previously missed using traditional keyword searching? Leveraging the AI-driven platform, users can more easily extrapolate the desired results with concept-based searches, rather than rely solely on specific, singular terms.
Micoli spoke further with LaFontaine at the closing reception. The two talked in detail about relevancy scoring and agreed that the IP industry is heading toward using AI to manage and search both patents and NPL. In addition, Micoli reminded LaFontaine that defensive publication is another strategy for protecting IP; IP.com’s machine learning algorithm combs through our extensive Prior Art Database of patents and NPL. This is always a third option, along with trade secrets and patenting.
After three days at the IPBC conference, it was clear the IP landscape is shifting. AI is no longer something from a sci-fi movie. Patent portfolio managers are required to gain more intelligence from the data. Investors are leaning toward algorithms. And, patent analytics software is providing a gateway to gain insight through data mining, machine learning, and artificial intelligence.