A few years into my budding career as an IPR&D information professional, I suddenly realized that I seemed to have forgotten how to search news articles, scientific literature, and other non-patent literature. This was especially surprising as before initiating my career in the intellectual property world, I was working as an R&D scientist in a chemical firm just after completing my doctoral research on catalysis.
During my lab research days, I would spend many of my waking hours trying to understand the state-of-the-art discoveries in related fields, identify the gaps or white-spaces, and establish the most fruitful opportunities for research. Physical journals, scientific papers and text books: almost all the content that I read and digested during my research days was non-patent literature.
The lesson to draw from this is: ignore non-patent sources at your peril – that might be exactly where you will find the nuggets of gold.
In which use cases do you need non-patent literature?
Non-patent literature is required for almost all use cases in patent search and innovation intelligence – except for freedom-to-operate/clearance searches and those related to IPR&D analytics which exclusively focuses on patent landscapes. For all other use cases, non-patent literature is a must.
It’s simple: as universities initially publish only a fraction of their research through patents, no prior art study or technology landscape is complete without looking at non-patent literature. And no technical competitive intelligence or innovation scouting project is complete without adding product information or technical literature.
Non-patent sources can be harder to find
At least in theory, all patent literature is now available in databases. In today’s world of full-text databases and ‘machine-translated text’, it is now easy to search for patents on a topic of interest and be sure that virtually all the patent literature can be searched and analyzed in an acceptably-short time frame, whereas in the past it would have been a gargantuan task.
So, to link it back to the concept of ‘recall’ – in theory a very high recall can be achieved for patent literature. Recall is the probability of finding all the relevant documents present in the public domain.
When it comes to non-patent literature or scientific articles, the situation is different. According to some accounts, the documentation of scientific discoveries/inventions started around the 17th Century. Unlike patents, the universe of non-patent or scientific literature is undefined primarily because of the extensive number of academic researchers writing and publishing their articles. Further, different formats (journals, books, conference proceedings, language, etc.) have made non-patent literature virtually impossible to search efficiently.
There have been multiple and ongoing attempts to capture, digitize and store this information. It’s a dynamic and complex task given the fact that integrating the data held on all the relevant databases would not be able to completely capture all scientific articles as the interests of publishers compete.
Now, while for patents all relevant documents theoretically can be found (recall close to 100% possible) this is not possible in non-patent literature. Even with a very-high effort invested, recall will max out at 70% in most technical fields.
Non-patent literature is easier to understand
The typical reaction of a young academic researcher when seeing a patent is to assume it is irrelevant without understanding it! Firstly, the claim language of patent literature is incomprehensible to the untrained eye. Secondly, complete texts of non-patent resources such as scientific articles are freely available, and this combined with their simplistic technical language is a great combination for a researcher!
After developing a formative understanding of what patents are all about, the types of claims, and so on, most IP professionals become proficient in the art of searching and analyzing patent literature for hundreds of hours on end: the more (sources) the better. This is of course true for data science in general, but perhaps the risks related to incompleteness are greater in the IPR&D world.
Panning for gold as the river flows ever faster, ever deeper
Today, patent-literature searching can be easier than non-patent literature searching for one very good reason: there’s a defined and structured universe – the patent offices – where the total data can be found, mainly in a digitized form. That said, even the patent universe keeps expanding as more patents are added. In addition, multiple patent databases are available that capture in one place the patents published by multiple patent offices. So, the trick lies in having the right combination of skills to a) understand the technology to be searched, b) search for patents, and c) interpret with right context (which depends on the objective of study).
Non-patent sources in comparison are virtually limitless – and ever growing. Just relying on patent searching would be like covering only a quarter of available prior art. So, ignoring this precious information source can have serious and damaging consequences for researchers and companies, as ‘re-inventing the wheel’ or not using the existing scientific know-how can lead to a significant waste of time and money – both on the actual scientific research and the ability to seek the right patent protection, because as they should say: ‘Ignorance is not bliss’…
Tell us about your experience!
The Information Adventurers blog is designed to be a source of information and experience on IPR&D searching. We’d love you to share some of your stories and experience with the community on our posted topics – anonymous or otherwise – so that we can all improve the quality of the work that we do. As always, questions and comments are also much appreciated!