Patent Class Searching - The Right Way

Patent searchers seem to be fairly opinionated about class searching. Some think it is essential, while others avoid it like the plague. I used to lean towards the “plague” side, on the theory that I could never be sure I knew all the relevant classes, so I’d rather stick to keywords.

That changed once I had access to facets. Facets make class searching a no-brainer. How? Read on…

Let’s assume I’m doing a patentability search for new titanium alloys that are particularly appropriate for bicycle frames. My query is ‘titanium bicycle frame’Note that I am purposefully leaving this search quite broad. I haven’t included terms like “alloy” or related terms such as “vanadium” or “silicon” (common alloying elements for titanium). I haven’t tried to refine it in any manner. I want a very broad search so as not to miss anything.

Running that search across all collections, I get almost 5,000 results. Too many to review for most searches.

That is where classes come in: CobaltIP can facet (filter) on classes, and the class definitions pop-up for you to instantly decide if they are relevant or not.

This allows me, within seconds, to note some interesting things about the search results. For example, Class 514 has the most documents in the search results, but Class 514 is about Drugs. Definitely not relevant. In fact, as I scan down the list, I note many classes that are not relevant to the search.

Since I can scan down the list and quickly see which classes are not relevant using the pop-up definitions, all I have to do to narrow the search, with very little risk, is to select “Check All,” and then uncheck the classes that are obviously irrelevant.

After unchecking the irrelevant ones, I’ve gone from 39 classes to 10 classes, and from just under 5,000 documents to 818 documents.

Within minutes I have a reasonably-scoped search, I’ve used very broad keywords, and (this is key) I’ve only filtered out classes that are obviously irrelevant.

Why is that last statement so important? Because it is the opposite of what most people do when class searching. Most people try to make a list of classes they want to include. With facets, you just take the whole list of classes from your base search and exclude the ones you do not want. That way, you know you haven’t missed a relevant class.

This example had a total of 39 classes, with 10 of them potentially-relevant. Maybe you could have developed that same list of 10 relevant classes by hand. But, why risk it? Also, note that some searches have far more relevant classes. In situations where there are dozens and dozens of relevant classes, you can never be sure you’ve thought of every one. So don’t try. This way is far easier, faster, and safer.