This site uses modern web standards that aren't supported by your browser. For best results, please upgrade to Google Chrome, Microsoft Edge, or Mozilla Firefox.


Importance of Name Normalization in Patent Research

|

IBM has been a market leader for decades. At one time, it seemed every office featured at least one piece of IBM equipment and computers were defined by whether they were “IBM compatible.” The century-old firm has continue to innovate to transition to a company that focuses primarily on services and software holding its status as a patent powerhouse; in 2014, IBM again ranked first among companies issued U.S. patents, just as it has for more than 20 years.

Surprisingly for a company as well known as IBM, anyone who wants to peruse IBM’s extensive patent library or research its intellectual property holdings, however, is likely to run into a significant hurdle—the company’s name.

Is it IBM? International Business Machines? IBM International Group? IBM Coroporation [sic]?

According to patent authority records, the answer is “all of the above.” And actually, those examples comprise just a small sampling; there are several dozen other variations of the company’s name listed as the assignee among the tens of thousands of patents the company holds.

Name variations such as these on official documents are part of the patent business. Similar fragmentation of assignee name afflicts the records of most companies with extensive patent holdings, particularly those that have been around for a while, as IBM has.

Each time an IP attorney or patent agent filling out patent application paperwork misspelled the word “business,” and each time a data entry clerk or automated document scanner eliminated a space between words, and each time any other comparable error was introduced, a new name variation was born.

Tackling the patent research issues created by assignee name fragmentation was the goal of a project to standardize and normalize assignee name data within LexisNexis® TotalPatent®.

First, assignee names were standardized using an algorithm that eliminates case variations, punctuation anomalies and obvious typographical errors. It also renders languages such as Chinese or Korean into Latin-based characters.  This step can winnow down the most severely fragmented assignee names by a factor of 10 or more, greatly improving the accuracy of search results. 

But that wasn’t the end of the process. The next step normalizes the names so that the likely variants of, say, IBM will be returned by a search for “IBM.” This project, which must be implemented manually, is ongoing. While not all of the assignee name data has been normalized, assignees with the IP significance and portfolio size of IBM typically have been processed.

If a researcher wishes to expand the search beyond the normalized fields, options remain to search assignees by either the original, “as published” entry in the USPTO records or by the standardized, pre-normalized form. As with any TotalPatent data, the search results can be exported into numerous standard formats, including PDF and Microsoft Excel. That allows for easy importation into a PatentAdvisor briefcase or other portfolio analysis platform.

Name normalization and standardization is a great improvement for patent-related research. It makes everything from prior art searches to portfolio evaluations more efficient and more accurate, and it is just one of the ways that we continue to ensure that TotalPatent remains one of the most powerful tools in your IP arsenal.



Learn about LexisNexis TotalPatent