In the midst of this article IPRally discussed their appearance in the ML4Patents.com webinar series and include comments made during the session.
Then, the last week of October, we had the great opportunity to present at the ML4Patents webinar, hosted by Tony Trippe. The concept of this series of webinars is to focus on how machine learning and artificial intelligence can assist in pushing the envelope for patent search. If you missed it, our session can be watched here.
The session focused on the unique methodology and technology behind IPRally and Graph Neural Networks, the actual search process and a test of accuracy by classifying documents in a pre-defined data set (which is the same for everyone that is taking part in this series). It was clear that our approach is indeed unique, and that it produces very impressive results. Some of the unique benefits of IPRally were highlighted – namely transparency, explainability and control – and concisely summarized by Tony when he stated that “there is no way anyone can call this a black box” and ”This is fascinating (…) you can’t ask for anything more explicit, in the ability to pretty quickly judge and adjust the sensitivity... so this is very impressive”. We, and our customers, knew this already, but it was great to get the opportunity to display this to a wider audience of experts, as the lack of control and explainability has been one of the key obstacles for many in their journey towards trusting and adopting AI. Up until now, that is.
Read the full article at the link below.
©Copyright ML4Patents | Powered By Patinformatics