Jumping Fish

Catching Patents (101)

I previously provided a quick overview of patenting, and specifically of patentability, freedom-to-operate (FTO), and the importance of searching — especially patent searching — in both of those processes.  As I pointed out, in determining whether or not you pass the legal barriers to getting a patent (patentability), a government patent office such as the USPTO is going to search the “art” out there, including literature references, published and issued patents, etc., to see if your invention is (among other things) “novel” and “non-obvious” in light of that art.  And as I also pointed out, understanding your FTO requires an understanding of the issued and also published patents in whatever country or countries you’re planning to make, sell or use your particular commercial process.  And, as I finally noted, you should treat everything I said as only a preliminary introduction to a complex process, and you should seek out a patent professional’s opinion before making any irrevocable decisions that might effect your putative patent rights or your possible infringement of the patent rights of others.1

In light of the above, it should be abundantly clear that searching — including but not limited to patent document searching — is critical to patentability and FTO issues.  Since the title of this post is “Catching Patents (101),” you can and should reasonably conclude that what I’ll be talking about here is an introduction to patent searching (the “101”), but that this is only an introduction.  Since this is an introduction, let’s do the 101 thing and start with a concrete example.

Keyword Searches

A concrete example?  Yes.  Literally.  No, really, I mean literally.  How literally?  Go to the USPTO’s search site for issued patents (http://patft.uspto.gov/netahtml/PTO/search-bool.html), and type the word “concrete” in the TERM1 box and hit “search,” that’s how literally.  As of this moment that search retrieved … 121,597 issued patents; that’s a lot of patents, although given a) there are currently over 8 million issued US patents, and b) concrete is probably a pretty important area for patenting, it’s perhaps not as amazing a number as you might think.  By contrast, a similar search, but for the term “aardvark” turned up (as of this moment) only 118 issued patents, very few of which seem to be directly related to the animal.  US6,892,622, for example, is titled “Anti-mine Unit,” and refers to aardvarks only in the context of the company “Aardvark Clear Mine Limited;” not very interesting, but then again I wouldn’t have expected to find the aardvark — or Orycteropus afer for the taxonomically inclined — to be a hotbed of patenting activities.

So much for a first search.  We learned that concrete is hot and aarvarks are not.  Anything else?  Well, actually, a number of things.  That aardvark search — a quick scan down the first 20 or so entries doesn’t indicate any aardvarks the animal, which means that “aardvark” quite possibly isn’t a very good keyword if you want animal aardvarks as opposed to all the oblique references that we actually got.  Let’s try “Orycteropus afer” and, when that doesn’t work (it didn’t for me anyway), let’s just try “orycteropus.”  Still nothing.  Perhaps there are no aardvark the animal patents at all.  Or, to go back to the first post, perhaps we’re still not using a well designed search net to find them.

For the moment we’ll leave the question of whether aardvarks the animal exist in any of the patent databases; the important point here is the dilemma of knowing whether we found nothing because there is nothing to find; or, instead, whether we found nothing because we didn’t do a very good search.  Or, to be more specific, whether we did a good keyword search, because that is, after all, the kind of search we performed, and it’s important to understand that there are other kinds of searches that can be done.

What other kinds of searches?  Well, in addition to keyword searches, there are classification searches and what I’ll refer to as ambient knowledge searches.  These two topics are important enough that I’ll spend the rest of this post briefly discussing them.

Classification Searches

In this age of search engines it’s hardly surprising that the kind of searching we first think of is keyword searching, with “aardvark,” “Orycteropus afer” and “orycteropus” being three examples of keywords.  However, surprising as this may seem, at one time there were no computers; and, even after that, for a good long time computers weren’t ubiquitously used for searching.  But on the other hand patent systems have been around for a long time — the US patent system was started by George Washington on April 10, 1790 for example,2 and according to Wikipedia patents date back to 500 BC, in, of course, ancient Greece.3  So up until recently people weren’t searching based on keywords; instead, they were searching based on systems of classification.

When I was at University, I made various futile attempts to create my own systems of classification;4 by graduate school I had pretty much given that sort of thing up.  The same cannot however be said for the various patent offices; the International Patent Classification (IPC) system, for example, has about 70,000 categories,5 while the USPTO has 450 classes and 150,000 subclasses.6  There are a number of helpful guides to using these classification systems,7 but a simple hands-on example will capture the essence of the process.

To that end, go to www.uspto.gov/web/patents/classification/ and, in box C (“Search USPTO”), with “look in” left at “Patent Classification,” search for the word “concrete.” While over time the results may vary, as of this writing this search brings up two hits. Click on the first one and search for occurrences of the word “concrete” and you’ll see variuos hits which give you a sense of classification — but on the other hand things still seem a bit discordant.

That’s fine, no one said classification system searching was instantaneously illuminating, so go back and re-run the same search, but this time with “look in” set to “Index to Classification.”  Click on the first hit, and search for the word “concrete,” and you should find an entire index category for concrete, with a lovely conceptual index depending from that term, a whole cornucopia of categories that give insight into how to find various patent references in the database that might never have come up via keyword searching.8

Ambient Knowledge Searches

I have to admit it: I don’t use classification searches nearly as much as I ought.  The ought part is straightforward enough: not only are such searches highly illuminating of possible patents and patent publications that might be relevant to patentability/FTO,9 but patent offices use their classification systems in determining whether there is patent art that makes an application for patent not-novel, obvious, etc., and so understanding how various patent offices do searching will help you in getting some estimation on the kinds of rejections you’re likely to get if/when you apply for a patent.

But I’m fundamentally lazy; or, perhaps, I tend to work in IP areas where classification isn’t as useful.  Thus one of my specialty areas is biotechnology, and my own feeling is that cutting-edge biotech inventions tend to fall outside the logical confines provided by classification.  Whatever the reason, I don’t use the classification systems as much as I should.

Or at least I don’t at first — instead, what I do use is ambient knowledge searching and keyword searching to define up a corpus of information that, with luck, gives me a sense of what classification(s) I might look in.  So how does all that work?

Ambient knowledge is the term I use to describe the jumble of information about a particular subject area that we gradually accrue by reading about that subject, thinking about that subject, researching that subject, picking through stuff about that subject.  It’s not a methodical collection of information; but the information grows nevertheless.  And the resulting pile of knowledge provides a huge entry point into designing patent searches.

So again let’s give an example.  Let’s say we suddenly realize we misremembered the name of our favorite “A” animal — it’s not, after some reflection, “aardvark” but instead “armadillo.”  And this sudden insight came when watching “Arthur,” and we realized that, much as we love the Reed family of Aardvarks, those weren’t in fact the animals that our favorite aunt has stuffed copies of all over her bookshelves.  Ambient knowledge strikes.

We search the USTPO classification index for “armadillo” and find … nothing.  And then we do a keyword search for “armadillo” and find almost 1200 hits, some of which are to stuffed toys but some of which are pretty clearly to the actual animals. Now why should aardvarks warrant little if any scientific study when so much attention is apparently paid to the biology of armadillos?  Are armadillos charismatic megafauna when aardvarks are not?10

Ambient knowledge to the rescue.  Years ago I came across a fascinating article on how armadillos are one of the few animals that can be used to study leprosy.11  So the likelihood is that armadillo biology is studied because these animals are a model system for a horrible human disease, and I could focus my search on that topic by searching for both “armadillo” and “leprosy.”

Furthermore, with better recollection I might actually remember the names of the people doing seminal work with that model system, and I could use that ambient knowledge to search just for patents to those people.  And, having done so, I could look at the cover page of each of those patent applications or issued patents — where the classification numbers for those documents are listed — and see if they have any classification numbers in common.  And any of those non-keyword searches would very likely add considerably to my ambient knowledge of the subject, making my searching better, and thereby giving me better insight into patentability (if I’m doing leprosy work for example) or FTO (if I want to see if there are any patents out there on armadillos and leprosy that I need to be wary of for infringement purposes).

So in a nutshell catching patents is as important as catching fish; and keywords aren’t the only way to troll the patent document seas.  The topic is important enough to warrant a real example — as interesting as aardvarks and armadillos are.  Stay tuned for the next post, where I’ll delve briefly into an actual USPTO search, again on the principle that examples are worth more than exegesis.

  1. And also, please do read the Disclaimer []
  2. See www.uspto.gov/news/pr/2002/02-26.jsp []
  3. See en.wikipedia.org/wiki/Patent []
  4. I distinctly remember attempting to do that sort of thing in my organic chemistry classes with various ketones, aldehydes, ethers, esters and so on; the results were spectacularly unsuccessful. []
  5. See www..wipo.int/classifications/ipc/en/ []
  6. See /www.uspto.gov/patents/resources/classification/ []
  7. See, e.g., www.uspto.gov/web/patents/classification/help.htm []
  8. The index class/subclass entries are themselves clickable, and will take you to pages where you can bring up all the currently published or issued US patents with that class, subclass etc. []
  9. And I should note that, although this discussion is USPTO-centric, classification searches in Rest-of-World IPC classifications are just as useful. []
  10. A term coined by Richard Lewontin; see, e.g., en.wikipedia.org/wiki/Charismatic_megafauna []
  11. My ambient knowledge is much more the beachcomber’s accumulation of random fragments of shells and crab claws and driftglass and other bright shiny objects than it is coherent or consistent.  And what I find “fascinating” is probably equally idiosyncratic.  Well, in the words of the famous sailor, “I yam what I yam.” []