Keyword Patent Searching Part 2

Have you ever found yourself performing an advanced patent search and left scratching your head asking, “How in the world did this patent end up in my data-set?” The English language has over 1,000,000 words, but we still find patents using the same words to describe completely disparate technologies.

Take this for example:

Within aviation technologies, the term ‘plane’ means one thing, but within CAD (Computer Aided Drafting) technologies  the term ‘plane’ means something entirely different.  So how do we avoid wasting time reading patents and performing patent analysis on documents that have nothing to do with your core technology application?  Sometimes it is unavoidable and we must use patent tools and patent analysis techniques as our answer, specifically the use of NOT patent keyword operators.  When using the NOT patent search operator it must begin with an affirmative, I use the following as an affirmative to capture all patents (applications & grants) for my refinement:

(inno_patent_grant OR inno_patent_application) NOT (Cad)

Disclaimer:  Be very careful when NOT patent search operators, sometimes they can have unintended consequences and can remove patents that are relevant to your technology of interest.  I recommend starting by using the NOT patent search operator for the title field of the patent first, and then gradually expanding the patent fields broader and broader while keeping an eye on types of patents or company’s patents that are being removed by the NOT operator patent search string.

Title, Abstract, and Claims Oh My!

Determining which patents fields to search can often be a struggle and a challenge when performing advanced patent searches and can leave you feeling a little like Goldilocks.  @body is too broad, @abstract is too narrow,  but @(abstract,claims,title) is just right!  Take this example about how the number of patents matching a patent search string describing coffee terms can change drastically based on the field your are searching in your patent search string.

My personal preference is to cast my advanced patent search wide and then sort, refine, and prioritize using Innography’s Advanced Analysis.  This approach minimizes the chance of having a patent of interest slip through the cracks just because it doesn’t mention a particular term in the claims.  

Double & Triple Check

As much as I would like to discount the advice I was given of “go read a couple hundred patents” when I began my patent search education, there is some value in this approach.  Understanding how patents within your technology are written and how their claims are structured can be incredibly insightful and helpful in improving your patent keyword searches.  Take this example below:

Within Innography’s Advanced Analysis, you can easily see the top companies based on your search string with a simple click of the display options.  When observing the top companies within the patent tool, be sure ask yourself the following:

Are there companies that look out of place? If so, click to view their patents within the overview page and use the keyword highlighting function to see how and where the keywords are used throughout patent document. 

Are you missing key patents within your own or a competitor’s portfolio?

If this is the case, save the patents you are catching in your keyword string to a project and label as ‘keyword’, then perform a company or patent number search, save and label these patents as ‘company’.  Move to your project space, select your project and then perform search @label (company NOT keyword) within the box in order to identify which patents were not captured by your patent search string.  Select these patents and analyze in a new window use Innography’s Advanced Analysis Text Cluster to identify new synonyms or if the list is relatively short, using patent overview page or quick view to quickly access how and if missed patents describe your technology of interest.

Other ways to check your keyword patent search strings is by sorting and grouping by patent classification code (IPC, CPC, USC).  Using the pie chart visualization and the table grid view you can quickly access if there are technology definitions that look out of place. If you wish to remove patents within non-relevant classification codes navigated to your refinement panel, find the out of place classification code from the refinement panel and alt click to remove all patents within the non-relevant classification code from your data-set.

Keep your Enemies Close

Using Innography’s company search and company and keyword search alerts, you can keep a close eye on your competitors’ patent filings.  By keeping abreast of your competitors’ patent filing activity, you can identify new keywords and synonyms to enhance and refine your patent search strings based on the evolving lexicon in your competitors’ patents.

Finally, stay tuned for our next blog topic on semantic patent searching.  Remember the best patent research strategy is a multi-layered approach and with Innography Advanced Analysis patent tools you can easily perform keyword and semantic patent searches for your IP analysis and intelligence.

Leave a Comment