Do you know how to measure quality of a patent search or landscape? In this article Urs Dommann shares his experiences using the Search Quality Index (SQI) and explains how digital technology is improving search quality.
Urs Dommann, who heads up IP and R&D Solutions at Evalueserve, is passionate about improving knowledge processes and data science to optimize quality in patent search, and is a regular speaker on this topic at IP Service World. He enjoys sharing his expertise and insights about IP and R&D, and starting conversations as a basis of exploring – together with clients – new paths to success.
Urs created the Information Adventures blog after recognizing that moving through the complex world of data and information in IP or R&D is like venturing high into the mountains or other seemingly uncharted territory – something he loves to do in the Swiss Alps near Zurich where he lives.
Quantify your Patent Quality
As an industry, should we really accept as the norm a failure to quantify the quality of patent search and patent analytics? That we should be doing it is clear: there are significant legal and business risks relating to the decisions that companies make every day based on patent search or patent analytics. Multiple, expensive and unrepeatable parallel searches are not a long-term, sustainable approach to mitigating this risk. CEOs and other senior managers expect to make their $100-million decisions based on reliable facts. It’s our job to justify that reliability, so we need to be able to measure the quality (or accuracy) of a patent search or landscape.
“As an industry, should we really accept as the norm a failure to quantify quality of patent search and patent analytics?”
Quality of Patent Search and Patent Analytics Can and Should be Quantified and Controlled
How can we measure the objective quality of a prior art search, a patent or innovation landscape, a patent watch, or a freedom to operate investigation? Our clients worldwide want to understand this, so at my session at the IP Service World (on November 27 at 1.30 pm) I will be sharing what Evalueserve has learned in its quest to make its searches the best in the market.
My team at Evalueserve has recently started an extensive blog series, where we not only discuss current best practices, but also some anecdotes from our market experience – of the perfect search, the perfect searcher, and more. We have also begun to explore new ways of thinking and how to improve our processes based on our Search Quality Index (SQI) formula, and we share some guidance in using it:
with RA,E = Recall actual and expected, P A,E = Precision actual and expected and I A,E = Insights actual and expected.
Using the Search Quality Index (SQI)
The SQI – and its derivatives IQI (Intelligence Quality Index) and AQI (Alert Quality Index) – can effectively be used to decide if the use case is better served by human experts, tools, or a mixture of both – we think of it as combining Mind+Machine. The SQI should be used as basis for planning and assessing how to improve methodologies and identify success criteria. It is also a framework for discussion between the information professional and end user (e. g. a patent attorney) to qualify the search results.
Using digital technology and AI effectively
Understanding search quality is one thing, but improving it is the real goal. In my session at the IP Services World, I will demonstrate how digital technology, AI and AI-enabled workflows can help information professionals to run reliable, transparent high-quality search and Innovation Intelligence processes. Using SQI / IQI / AQI we have identified a clear definition of use case, resource efficient control of recall, smart data re-use and “data-driven insights” as key factors.
Evalueserve is the largest IP and R&D service provider; offering a unique approach to quality patent searching, patent analytics and innovation intelligence. Our revolutionary Search and Intelligence quality index allows, for the first time, the quality of a search to be determined. Coupled with customized curated data for better insights we bring a unique proposition to the market. We additionally offer industrialized, high volume digital patent drafting, prosecution and IP docketing. Our mind+machine™ approach is a unique combination of human expertise and best-in-class proprietary technologies (such as Insightloupe) which use smart algorithms to simplify key tasks, control and improve output of complex analytics task and provide an advanced digital experience to clients.