Assignment Best Practices

From Intellogist

Jump to: navigation, search
If you found this page through a web search, we invite you to visit our Main Page to see what Intellogist is all about.This Best Practices article exists for the community to share common wisdom and valuable insights about the prior art searching process. Registered users can add, edit, or delete material on this page. Information on this page is the result of community collaboration and, as such, is vetted by the community at large, not individual experts or fact-checkers. All information contributed to this page is public information - do not post confidential information. For more information about creating and editing Best Practices articles, please see our Help pages.


Contents


Introduction

Note: Some of the examples in this article are taken from the presentation "Finding Patent Assignees" by Ron Kaminecki of Dialog. Many thanks to Ron for his permission to use these examples, which are cited to him individually, below.

There are several major obstacles that can prevent searchers from finding documents that are assigned to a given corporation of interest. This article presents a number of scenarios that can prevent searchers from finding the entire patent portfolio of a given company, and solutions to help avoid these roadblocks. Note that even after following these steps, it is impossible to be certain that every single relevant patent document has been found.


Scenario 1: Missing Assignment Data

The assignment data is not printed in the assignment field of the document. Often assignee data is withheld and not printed on the published application face in the assignee field – this can happen in multiple national and regional patent systems, including the United States (US) system. In addition, it sometimes happens that patents may be issued without the appropriate assignment data published on the patent face, so they appear to be assigned to individual inventors when actually they are the property of a corporation. Finding these "hidden" documents can be critical to a search project.


Strategies

Search the Agent/Correspondent field:
Sometimes the name of the company will be in the "representative," "agent," or "correspondent" field of the patent. As a hypothetical example, a published application owned by Acme Inc. may not have any formal assignee information printed on the face of the document, however the attorney's (agent's) address may be written as "Acme Legal Department" or similar. Searching in the "attorney/agent," "correspondent," or "representative" field of your search service can retrieve these documents. PatBase, QPAT, MicroPatent PatentWeb, Thomson Innovation and other systems may index this field – TotalPatent does not appear to.

Search the assignment data which appears in various patent legal status databases:
Sometimes this data will not appear on a published application's face, but will be included in that country's re-assignment records. The following options describe these resources, most of which specifically focus on US re-assignments.

  • For US documents, search in the USPTO assignment database. The database on the USPTO website, at http://assignments.uspto.gov/assignments/?db=pat, is the primary source for US re-assignment data. For maximum safety, users should go straight to the primary source at the USPTO whenever possible. Or, if your necessary workflow is not supported there, you have additional options:
  • Determine whether your regular patent search engine incorporates USPTO re-assignment data. For example, PatBase includes USPTO re-assignment data, and searches it along with regular assignment data (both types of assignment data are contained in the assignee field). In addition, Thomson Innovation allows users to search this information via the "US Re-assignment Data" field. Note: there could possibly be some issues with timeliness here, as the update timetables for this re-assignment data are generally not disclosed publicly. It’s best that you check with your information provider to determine the currency of the data at any given point in time.
  • Search the IFI Claims database, which contains USPTO assignment data (updated on a weekly basis), as well as probable assignee information generated by IFI's in-house methods (see "Search a database that predicts probable owners", below). Services that would allow you to search this database would include Dialog Classic (or Classic Web), Questel on Imagination, or STN.
  • For both US and non-US documents, search INPADOC legal status data. If you have a service which allows you to actually search assignee names in the INPADOC legal status records (not just to display this data), you can query legal status information, which sometimes includes re-assignment information (on a country-by-country basis). This data has been provided to the EPO by multiple national and regional patent offices, and accuracy and timeliness cannot be guaranteed. A service that would allow you to search the INPADOC legal status data would be Dialog Classic (or Classic Web), Questel on Imagination, or STN. To verify whether a certain country of interest might have provided the relevant data to the EPO, see coverage information from the legal status file at http://www.epo.org/patents/patent-information/raw-data/useful-tables.html.
  • For individual countries, it may sometimes be possible to go to the national patent office's legal status register and search by assignee name – whether this is possible for any specific country must be determined by experiment.


Search a database organized by patent family:
Even if the search is for US patent documents only, expand to other non-US collections rather than limiting your search to US results. You may find non-US family members held by the assignee of interest, and these may allow you to infer that the company might hold the US family member as well. Running your search in the Derwent World Patents Index, or in any system that searches INPADOC family data, will probably be successful. However, the results of this kind of search must be carefully analyzed - keep in mind that two different companies could potentially hold the rights to the same invention in two different countries. In addition, when searching the INPADOC database, keep in mind that the company name could have been transliterated when the patent was granted in another language, and these results could be easily missed by a search on the company's English name (see Translations and Transliterations for more information).

Search a database that predicts probable owners:
Some patent databases use special logic to predict probable assignees for unassigned patents – for example, algorithms can check the representative/agent field to look for a corporate address, or may check the inventor field to see if a well-known corporate inventor suddenly has a patent application upon which his or her company name does not appear. Innography is one tool that uses a proprietary method to predict probable ownership. Another database, IFI Claims, is a pay-per-use database available on Dialog, Questel and STN that contains predicted assignees. Since they both use their own methods, the same query run in these two databases may come up with different results, and it may be useful to query both if an exhaustive search is needed.


Scenario 2: Multiple Name Variants

The company name has multiple variants, and/or has been transliterated to or from a language that does not use Latin characters. This concern will affect EVERY assignee and inventor name. The reality is that no company name can totally escape from misspellings when it is represented on patent documents, and that any patent which is filed internationally may have the company and inventor names translated into other languages, and perhaps transliterated back into English.[1]


Background Information: Multiple Name Variants

As an example of this problem, consider the company DuPont, the full formal name of which is "E I DuPont de Nemours." The PatBase "Browse Index" function shows the following name variations under "DuPont":

"DUPONT" OR "DUPONT_AUSTRALIA_LIMITED" OR "DUPONT_AUSTRALIA_LTD" OR "DUPONT_E_I_DE_NEMOURS" OR "DUPONT_E_I_DE_NEMOURS_AND_COMPANY" OR "DUPONT_S_T_" OR "DUPONT_U_K_LIMITED" OR "DUPONT_U_K_LTD" OR "DUPONT_UK_LIMITED"

And it shows even more under "E I DUPONT De NEMOURS":

"E_I_*_DU_PONT_DE_NEMOURS_AND_CO" OR "E_I_*_DU_PONT_DE_NEMOURS_AND_COMPANY" OR "E_I_ADUAPONTADEANEMOURSAANDACOMPANY" OR "E_I_D_PONT_DE_NEMOURS_AND_CO" OR "E_I_D_U_POINT_DE_NEMOURS_AND_CO" OR "E_I_DD_PONT_DE_NEMOURS_AND_COMPA" OR "E_I_DE_NEMOURS_AND_CO" OR "E_I_DE_NEMOURS_AND_COMPANY"

As you can see, even a broad search on the keyword "DuPont" would not necessarily bring up the patent documents assigned under the names in the second string. Although searching a patent database organized by family, such as PatBase, QPat's FamPat, or the Derwent World Patents Index (DWPI), may ameliorate the problem somewhat by bringing family member records together (so that, for example, a document with an erroneous company name might get the correct assignee name associated with it via a family organization scheme), it is still absolutely vital to perform these index checks. Without them, a comprehensive search on an assignee name cannot take place.

In addition, company name variants can also be created from some company's patents that are filed in Japan. The classic example is that Japanese company names are often followed by the words "Kabushiki Kaisha," often abbreviated as "KK" (especially in files with standardized assignee names, such as the DWPI.) In PatBase, the company Fuji Heavy Industries has the following entries (this is an incomplete list):

"FUJI_HEAVY_INDUSTRIES_LTD_JA" OR "FUJI_HEAVY_INDUSTRIES_LTD_TOKIO" OR "FUJI_HEAVY_INDUSTRIES_LTD_TOKIO_TOKYO" OR "FUJI_HEAVY_INDUSTRIES_LTD_TOKIO_TOKYO_JP" OR "FUJI_HEAVY_INDUSTRIES_LTD_TOKYO_JP" OR "FUJI_HEAVY_INDUSTRY_CO_LTD_JA" OR "FUJI_HEAVY_INDUSTRY_KABUSHIKI" OR "FUJI_HEAVY_INDUSTRY_KABUSHIKI_KAISHA" OR "FUJI_HEAVY_INDUSTRY_LTD" OR "FUJI_HEAVYINDUSTRIES_LTD"

In the DWPI file on Dialog, the list could look like this (an incomplete list):[1]

E6 6127 CK=FUJH (FUJI HEAVY IND LTD)
E7 1 CK=FUJH (FUJI HEAVY INDS KK)
E8 4 CK=FUJH (FUJI HEAVY INDS LTD)
E9 3 CK=FUJH (FUJI JUKOGOYO KK)
E10 1 CK=FUJH (FUJI JUKOGYO CORP KK)
E11 1725 CK=FUJH (FUJI JUKOGYO KK)
E12 3 CK=FUJH (FUJI JUKOGYO KL)
E13 2 CK=FUJH (FUJIJUKOGYO KK)

Note that the Derwent World Patents Index (DWPI) uses the same standardized four-letter code to represent variations of this company name: FUJH. Derwent can do this because they use a team of human indexers to examine patent documents, and they assign these codes to documents (from the top 4,000 companies by patent volume) to increase retrieval accuracy for assignee searching.


Background Information: Translations and Transliterations

As an example of the problems that can occur as a result of translations and transliterations, consider the following family members and their member-level bibliographic data entries, all from the same family record in INPADOC (notice the incorrect translations): [1]

UNITED STATES OF AMERICA (US)
Patent (No,Kind,Date): US 4130178 A 19781219
ELEVATING DEVICE (English)
Patent Assignee: SMITH JR RAYMOND E
Author (Inventor): SMITH JR RAYMOND E
Priority (No,Kind,Date): US 782183 A 19770328
Applic (No,Kind,Date): US 782183 A 19770328
National Class: * US 182141000
IPC: * E04G-001/22
Language of Document: English
ITALY (IT)
Patent (No,Kind,Date): IT 1091358 A 19850706
DISPOSITIVO ELEVATORE (Italian)
Patent Assignee: SMITH RAYMOND E JUN (US)
Priority (No,Kind,Date): US 782183 A 19770328
Applic (No,Kind,Date): IT 7769682 A 19771128
IPC: * B66F
Language of Document: Italian
JAPAN (JP)
Patent (No,Kind,Date): JP 53119556 A2 19781019
JACKING APPARATUS (English)
Patent Assignee: REIMONDO II SUMISU JIYUNIA
Author (Inventor): REIMONDO II SUMISU JIYUNIA
Priority (No,Kind,Date): US 782183 A 19770328
Applic (No,Kind,Date): JP 77152826 A 19771219
IPC: * B66F-003/22
Language of Document: Japanese
Patent (No,Kind,Date): JP 81041559 B4 19810929
Priority (No,Kind,Date): US 782183 A 19770328
Applic (No,Kind,Date): JP 77152826 A 19771219
IPC: * B66F-007/08; F16H-021/00
Language of Document: Japanese

Note that if you were searching the INPADOC file using the assignee name, any one of these name variations would have pulled up this record. However, if you were searching, for example, in a Japanese-only database without searchable family data, your search on Smith, Raymond would probably not have found this record.


Strategies

  • Use a search tool with a "browse index" or "expand" feature to find all name variants, and plan for those variants in your search string.
As an important side note, not all indexes are created equal. PatBase's index seems to retrieve terms alphabetically based on the first letter of the term of interest, while the Dialog index browsing feature (the "expand" command) treats each word in a phrase separately, so you pull up all relevant results, no matter where the term of interest appears in the actual indexed data.[1] For example, when using the browse feature in PatBase, you may need to try other arrangements of the phrase, such as both "DuPont" and "E I D…" to get the two strings in the example above. It is a good idea to become familiar with the browse index feature of the search product you are using.
  • Use a search tool that uses a computer algorithm to standardize assignee names.
    Innography is one example of a tool that does this – however, the danger is that the computer standardization is somewhat of a "black box" that can possibly include unwanted documents or exclude desirable ones. Innography's feature includes a suggestion form where human users can point out errors in the algorithm, however it may not be wise to use this process exclusively if you need a 100% complete and comprehensive results set.


Scenario 3: Subsidiary Names

The company may (and will likely) have subsidiaries that go under a different name. Subsidiaries are parts of a larger company that are branded with a different name than the parent company. Searchers need to ask the customer/search requester which subsidiaries might be of interest, rather than automatically including all subsidiary names.


Strategies

  • Use a search tool with a corporate hierarchy database. Examples of search tools that include a "corporate tree" or corporate hierarchy database are TotalPatent, which includes LexisNexis Directory of Corporate Affiliations (DCA) data, and Delphion/Thomson Innovation, which both include 1790 Analytics data. When possible, check both corporate data resources, since they draw from different data sources.
  • Perform web research or other research to learn more about the companies of interest, and talk to the person requesting the search.


Scenario 4: Re-assignments

The company may now own patents that were originally assigned to a different company. Before any assignment search can be complete, users need to consult databases with re-assignment information.


Strategies

Note: This information is also included in Scenario 1, above

  • For US documents, search in the USPTO assignment database. The database on the USPTO website, at http://assignments.uspto.gov/assignments/?db=pat, is the primary source for US re-assignment data. For maximum safety, users should go straight to the primary source at the USPTO whenever possible. Or, if your necessary workflow is not supported there, you have additional options:
  • Determine whether your regular patent search engine incorporates USPTO re-assignment data. For example, PatBase includes USPTO re-assignment data, and searches it along with regular assignment data (both types of assignment data are contained in the assignee field). In addition, Thomson Innovation allows users to search this information via the "US Reassignment Data" field. Note: there could possibly be some issues with timeliness here, as the update timetables for this re-assignment data are generally not disclosed publicly. It’s best that you check with your information provider to determine the currency of the data at any given point in time.
  • For both US and non-US documents, search INPADOC legal status data. If you have a service which allows you to actually search assignee names in the INPADOC legal status records (not just to display this data), you can query legal status information, which sometimes includes re-assignment information (on a country-by-country basis). This data has been provided to the EPO by multiple national and regional patent offices, and accuracy and timeliness cannot be guaranteed. A service that would allow you to search the INPADOC legal status data would be Dialog Classic (or Classic Web), Questel on Imagination, or STN. To verify whether a certain country of interest might have provided the relevant data to the EPO, see coverage information from the legal status file at http://www.epo.org/patents/patent-information/raw-data/useful-tables.html.
  • For individual countries, it may sometimes be possible to go to the National Patent Office's legal status register and search by assignee name – whether this is possible for any specific country must be determined by experiment.


Sources

  1. ↑ 1.0 1.1 1.2 1.3 Kaminecki, Ron. "Finding Patent Assignees." Presentation given at PIUG Northeast, 2008, and provided via e-mail correspondence, October 16, 2008.

Retrieved from "http://www.intellogist.com/wiki/Assignment_Best_Practices"
Patent search questions. Expert answers.  Brought to you by Landon IP
HOT Items

Intellogist is brought to you by the patent search experts at Landon IP.

Welcome to Intellogist!

To network with our international community of patent info pros, please create an account.

For a list of our current members, see our Community Page.

Navigation