Things discussed in this blog: How do Patent Examiners do research to determine if patents should be granted? Are Library clients mostly skilled or unskilled researchers? Will Semantic Searches replace Boolean Searches?
I’ve been thinking a lot about how patent examiners do their research. I’ve been doing brief literature searches on the topic the past couple days and have found barely anything. You would think with the “Patent Wars” going on between Tech Giants and the anti-intellectual property movement ramping up everywhere that there would be more in-depth evaluations and critiques of how patent examiners do their jobs. You have hundreds of tech writers arguing that Patent Examiners are assigning “bad patents” that stifle innovation and yet there are no substantive studies on the research process patent examiners go through in order to decide how to decide if something should be patentable?
Patent Examiners have a duty to do a substantial literature search to make sure that the main concept behind the Patent is novel, usable, and inventive. Particularly where the novel and inventive qualifications are concerned it is important that Patent Examiners do an in-depth literature search to make sure that this is an original idea for a patent.
Now note that I am saying that there is a lack of information on how Patent Examiners do their research and not that they do bad research. On the contrary, I have met some brilliant patent examiners with Phds and who are very aware of the developments in their field. Patent Examiners, while they are divided into subject sections, usually do inevitably come into contact with complex jargon that they may be unfamiliar with – but I have been nothing but impressed with how quickly they can pick it up and learn to apply it. Also after sitting in some database training sessions with them and watching them extract the key concepts from mind-numbingly boring and complex documents, I have no doubt that they are amazing researchers.
This is all my own anecdotal experience though and as a librarian it kind of makes sense that I would only interact with the best researchers while the other ones stay away. Individuals who have a drive to acquire new research methods are interested in learning about new things in research are probably more likely to seek out library help. …. I can hear your objection: “Don’t bad researchers need research help and so they would seek out library help while good researchers don’t need it so they would avoid it?” And yes. That is true. But research tools aren’t static things and so good researchers would come to new database training presentations. Also a large part of what a patent office library does is obtain specialised articles. Better researchers would make more use of interlibrary loans and need more specialized articles.
Ok. I just think I brought up an interesting question: Is a library used more by good researchers or is it used more by bad researchers? Answer is both. And a good library should help those bad researchers turn into good ones.
What I worry about is those average researchers. The ones good enough not to need the basic help libraries offer but not good enough to realize that they could use the advanced help and services libraries offer. There are a lot of these researchers. A lot of university students fit this category – so familiar with keyword searching from google and website organization that they can find basic information easily but ultimately no idea how to find journal articles about complex concepts or navigate a specialized database.
It is the Libraries duty to reach these average researchers. Here is where library marketing and training are important. To reach those researchers who think their skills are “good enough” and don’t try to improve them.
That was a long tangent. I think it was an important point to make though. Let’s go back to Patent Examiners information retrieval methods. A lot of databases now are pitching semantic or concept based search engines to patent examiners. Doing a search for information about a Patent requires a lot of Boolean keywords to find synonyms and similar concepts but Semantic searches do the “similar concept search” directly without needing Boolean. So instead of an Examiner needing to rack their brain and doing a “pre-search” before their official Boolean query to determine Boolean terms that can be used, they can just copy and paste a Patent Claim in the search box and boom. Search Results.
Do Semantic Searches work? Do we trust a machine to find synonyms more than a human? Who would make more mistakes? All good questions. You already see databases with “smart text searching” that do semantic searches for you automatically if Boolean searches don’t return any results. Also a lot of databases are combining “Semantic Searches” with Boolean queries in order to help organize results by relevance. My guess is that Boolean might be joining the Browse* search function in a few years.
Another tangent. Back to Patent Examiners. Examiners also have a huge emphasis on having a fast turnaround, which doesn’t give them time for indepth research. That is clearly a problem. Also there shockingly is this “one patent and one examiner” way of thinking for patents. You would think that multiple examiners would look over and discuss a patent. Though I am not 100% sure on the process here maybe Patent Examiners do talk amongst themselves for complex patents and share ideas. But this is the problem! We don’t know. Somebody needs to do a study on how Patent Examiners get information about whether they should grant a patent.
Ok. That’s about all I wanted to say. Please share any insights if you have any.
* “Whatever Happened to Browse?” Online Searcher 39, no. 3 (March-April, 2015): 71-73.