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.


Patent research: What’s in a name?

|

Have you ever tried to find all the patent applications filed by a single organization or tried to analyze a result set by applicant name? Undoubtedly many patent researchers have found enormous variety in the names used by some organizations, leading to doubts about the ability to uncover all prior art. In addition, many organizations—for instance, Asian-based companies that file with international IP offices using non-Latin characters—may show only in Chinese, Japanese, Korean or other Asian-character results sets.

Did you find them all?

The process
To overcome these problems, advanced patent research tools by LexisNexis® include a Normalized Assignee Name to:

  • Assign a Latin name to where only the non-Latin name has been provided
  • Automatically generates a ‘Standardized Name’ using an advanced algorithm
  • Manually generates a ‘Normalized Name’

The purpose of the standardization process is to create names in the same Latin-based font using generic characters, address misspellings, group name variations and remove obsolete information. The normalization process goes beyond this and manually groups all names that belong together, using well-known names in the patent community while taking into account all name variations and ensuring that any normalized name is unique.

Example
IPC: F21S2/00
Keywords: helium or oxygen or hydro* or neon or carbon*
Grants only

Results:

  • 2619 documents found, of which 4 contained no applicant name and 3 were identified as non-company names, leaving 2612 namesof which 851 are unique.
  • 2232 non-Latin names, of which 1036 are unique. 1967 (88%) Latin names could be assigned to the non-Latin names.
  • 2366 standardized names, of which 873 are unique.
  • 1254 normalized names, of which 104 are unique.

A simple analysis of the applicant name would look like this:

TP_Normalization2

Standardized name analytics results     
TP_Normalization1  
   
Normalized name analytics results  
TP_Normalization3  

Taking a precise look at Panasonic, we see that in the original data, 19 unique names are used, 4 with Latin characters, 15 with non-Latin characters. These unique names are reduced to 8 variations of the standardized name and to a single normalized name.

The volume of the LexisNexis patent database

Total Raw Company Names 39.8 million
Unique Raw Company Names 185 thousand
Unique Standardized Names 6.4 thousand
Unique Normalized Names 2 thousand

Almost three-quarters of all non-Latin names in the LexisNexis patent database are currently assigned a Latin name using name standardization or normalization algorithms. LexisNexis has created 2,000 normalized names—covering approximately 50 percent of all raw names—and continues to add more normalized names daily. Our process assures an accuracy of 99.8 percent.

What’s in it for the user?

  • Searchability: Already having completed a large amount of research on name variations, searches performed in a LexisNexis patent database will automatically provide all relevant variations—including those in non-Latin languages.
  • Analyses: Highly reliable name variation analyses have already been performed—many of which may not have even been considered yet. 
  • User experience: Obtain sets of comprehensive and accurate data while being assured that even the tiniest variations have been covered.
  • Efficiency: Save time by not having to build new queries to uncover all name variations.
  • Flexibility: Choose the preferred name (raw, standardized or normalized) to optimize the query and anticipated results.

 



Learn about LexisNexis TotalPatent