Video Transcripts : Patent Searching

Searching for patents can be challenging because patents often use complex language. Being strategic about keywords can help you find relevant patents more efficiently. To start, think about words that describe your idea and synonyms for those words.

Specifically, you can think through the function of your idea: What does it do? The composition and structure of your idea: What is it made of? And its intended use: What is it used for? For example, if you are trying to find the patent for the Slinky toy, you might brainstorm functions like goes down stairs or steps or a slant or incline, expands and contracts. For composition and structure, you might come up with words like coil or spiral and metal or plastic, and its intended use is as a toy for play or for entertainment.

Or if you're a bioengineer, you might be working on a project about how to adhere a stent graft to an artery. You might brainstorm keywords for functions like stent-graft coating or adhesive. It might be made of a polymer or hydrogel, and its intended use is for EVAR which stands for Endovascular Aneurysm Repair. Specifically, our idea is meant to prevent endoleaks.

When brainstorming keywords, keep in mind that you may need to break an invention down into its component pieces. For example, there is no one patent for Apple's iPhone. Instead, there are hundreds of patents covering aspects of individual components, such as the touch screen, microphone, Wi-Fi capability, processor, and so on.

Once you have some keywords that describe your idea, you can use AND, OR and NOT to connect those keywords to create an effective search. For example, this table shows the keywords we brainstormed for function, composition and structure, and intended use, and how we might connect them.

We can connect synonyms using OR. This tells the database we want our results to include either polymer or hydrogel. We can also use AND to connect the concepts. This tells the database we want results to include all of these concepts. We can then translate that into a string of keywords to search. We put each concept in parentheses connected by OR.

For example, coating or adhesive. We can use quotation marks around phrases such as "endovascular aneurysm repair". The quotation marks tell the database we want that specific phrase. Without the quotation marks, the search would look for those words in any order. Then we can connect each concept with AND.

Notice that our search includes an abbreviation EVAR, as well as the spelled out phrase. It can also be useful to include alternate spellings. For example, if stent-graft is sometimes spelled without a hyphen, we can include that version of the word as well. For a more detailed walkthrough of how to connect your keywords using AND, OR, and NOT, review the tutorial called How to Build an Effective Search.

Other tools may include different patents. For example, patents from different countries or different time periods. So for a complete patent search, it's important to use multiple databases. And different tools offer different filters to help you narrow your search.

For example, the patent database Lens.org offers some filters similar to Google Patents like date range and status, as well as other options that allow you to filter with more precision, like jurisdiction, applicants, inventors, owners, agents and attorneys as well as cited works, classifications and document family. For more information on using tool-specific filters and features, check out the library's other patent searching tutorials.

Once you've found a relevant patent, most patent databases provide ways to find other related patents. For example, Google Patents provides Find Prior Art and Similar buttons, and lists patent citations near the bottom of each patent.

Patent records also include the CPC classification, which stands for Cooperative Patent Classification. This CPC info is a good jumping off point for the other strategy you need to know for a complete patent search: Classification searching. For more information on how to do this, check out the next tutorial called Patent Searching: Tips for Using Classifications.

Searching by classification helps ensure you don't miss anything when you look for patents as part of your background research. Patents are organized into an international system of classifications or categories. Like Netflix categorizes shows into comedies, dramas and other genres, or how Amazon divides products into categories like clothing, books or games.

The classification system for patents is called Cooperative Patent Classifications or CPC for short. The first screenshot shows the high level CPC classifications. For example, Class A is human necessities. Class B is performing operations and transporting. C is chemistry and metallurgy and so on.

The second screenshot shows the subclassifications for H electricity. For example, H is a subcategory for basic electronic circuitry, which also has subcategories, including HB Generation of Oscillations and HC Modulation, and so on. So all patents related to Basic Electronic Circuitry, Modulation would be classified under HC.

That's why if you can find the relevant CPC classes for your project, you can more easily see what ideas have been patented related to your topic anywhere in the world, even if different terms have been used to describe them. You'll also likely have fewer irrelevant results to look through.

So how can you find classifications for your project? You can start by brainstorming keywords that describe your idea. For more information on how to generate useful keywords for patent searching, check out the tutorial Tips for Using Keywords.

Once you have your keywords, you can use those keywords to find relevant classifications using a few different strategies. You can use your keywords to search the list of CPC classifications. You can use your keywords to start searching for patents and see how relevant patents are classified. And you can browse an expandable list of classifications to find related classifications that may be relevant for your project. Let's look at each of these strategies.

The first strategy is to search for your keywords directly in a list of CPC classifications. Many patent searching tools offer searchable CPC lists. For example, on the Lens homepage, you can select show classifications to bring up a searchable list. Make sure CPC is selected. Then enter your keywords in the search box.

In this example, a search for stent grafts brings back a list of classifications with brief descriptions of each one. You can review related classifications, both broader and narrower, by selecting the folder icons to expand or contract the list. As you review classifications and their descriptions, you can select the ones you think may be relevant to your project using the checkboxes. Then select New Patent Search to see a list of patents within those classifications.

If you've already found some relevant patents, a useful strategy is to see how those patents are classified. Then you can search for those classifications. For example, if we find a relevant patent in Google Patents, we can view the list of classifications for that patent.

In this case, it seems like the sub classification AF/ is most relevant to our search, so we could select it to see a list of patents in that classification. If you use this strategy, keep in mind that your project may have more than one classification that is relevant. To make sure you're not missing any, keep a list of relevant classifications as you review patents and search or browse the CPC lists as well.

Another strategy is to browse the CPC list. For example, as we saw earlier, you can expand any category within the Lens classification list to see the classification title and description. You could use this list to drill down and explore categories that you think might include classifications that are relevant for your project.

Another patent database called Espacenet offers a full screen classification list that you may find a bit easier to browse. For example, you can click on Classification A Human Necessities and scroll through the relevant sub classifications. Similar to Lens, you can use this list to build a new search for patents in those classifications. This can be useful as a final step to make sure you're not missing any potentially relevant classifications.

As we learned in the last few examples, most patent databases enable you to search for classifications as you go. However, no search engine or database is guaranteed to show you everything that you might need for your project, because different databases include patents from different countries and time periods. So the best practice for a complete patent search is to search for classifications, jot down any classifications that seem relevant, and then search for those classifications in multiple patent databases.

For example, we can copy and paste the classification we found in Google Patents into the search box on Lens.org or any other patent database. Then we can review that list of results to see if there are any other patents we might need to include in our review. For more information about using different patent databases, please check out the library's other patent searching tutorials.

As you find relevant patents, don't forget to check for related patents. For example, using Google Patents' Find Prior Art and Similar buttons, and the citation network. It can also be useful to keep track of the patents you review in a list or a citation manager. Keeping a list as you go can help you avoid duplicating efforts and save time when you cite relevant patents in your project report.

Finally, remember that research is always an iterative process and this is especially true for patent searching. Try a search, evaluate your results, adjust your search, and try again.