In my last post, I discussed the approaches taken by four commercial search systems with regard to their Japanese collections: Minesoft PatBase [1], Questel’s QPAT [2] and orbit.com platforms, Thomson Reuters Thomson Innovation [3] , and LexisNexis TotalPatent [4].
This post contains a summary of the comparison information in table form. The tables below reference “machine-assisted” and “machine-aided” translations; for an in-depth explanation of what these terms mean, please see my earlier post in this series [5]. This post relates to searchable electronic text coverage; PDF image coverage is not presented here.
Before I get to the comparison, just a quick note: although I have created this post by summarizing publicly available information vendor system help files and, in some cases, statements made from vendor representatives, readers should verify coverage with providers before basing any purchasing decisions on this information.
Coverage for Japanese Published Applications
|
Collection type: Published Applications |
Minesoft PatBase |
Questel’s QPAT/orbit.com platforms |
Thomson Reuters Thomson Innovation |
LexisNexis TotalPatent |
|
Available text |
Full Text |
Full Text |
Full Text |
Full Text (only English abstracts are keyword searchable) |
|
Language (and translation type if applicable) |
Original Japanese language
(English hand translated abstracts available from 1976)
|
English “machine –assisted” translation
(English hand translated abstracts available from 1976)
|
English “machine –aided” translation
(English hand translated abstracts available from 1976)
|
Original Japanese Language
(English hand translated abstracts available from 1976)
|
|
Years of coverage |
From 1998
|
From 2004
|
From 2003
|
From 1991
|
|
Notes |
The collection is original language; however, on-the-fly machine translations into English are available. |
The collection is made up of Questel’s in-house “machine-aided” translations
Bibliographic coverage in FamPat family file extends back to 1973
Japanese assignee names are translated by hand |
The collection is made up of Thomson Reuters “machine-assisted” translations
|
Bibliographic coverage extends back to 1956 |
|
F-I/F-terms available? |
Yes |
Yes |
Yes |
No |
The next table presents a brief summary of other JP kind codes which are covered by the four databases in question. TotalPatent currently offers a collection of granted patents, while QPAT and orbit.com offer full text machine translations of utility models. Thomson Innovation offers full text collections for both granted patents and utility models.
Coverage for Japanese Granted Patents and Utility Models
|
Other Collections |
Minesoft PatBase |
Questel’s QPAT/orbit.com platforms |
Thomson Reuters Thomson Innovation |
LexisNexis TotalPatent |
|
Granted Patents – |
N/A |
N/A |
Full Text English MT
From 2005 (full);
Thomson Reuters “machine-assisted” translations |
Full Text Original language
From 2003; bibliographic from 1913
(only English abstracts are searchable)
|
|
Utility Model Applications |
N/A |
Full Text English MT From 2004
Questel “machine-aided” translations
|
Full text English MT From 2008; bibliographic from 1971
Thomson Reuters “machine-assisted” translations |
N/A |
Additional Notes from Vendors
In addition to the comparison table above, there are several other factors that could influence a decision about which service offers the best coverage to suit a user’s needs. Below, I’ve compiled some of these additional statements, which have come from the system’s help documentation and in some cases, system representatives.
- Questel’s QPAT and orbit.com platforms: Currently, the time delay from when new JP documents are published to when they appear in the Questel database is as short as 1 week. Questel representatives have stated that, thanks to the company’s exclusive relationship with the Patolis corp [6], they are able to provide Japanese F-I and F-term classifications in both their JP full text files, and FamPat family database. It’s also worth noting that for an extra fee, Questel provides access to Japanese legal status information from Patolis, which will continue even as Patolis undergoes a bankruptcy and re-organization. Patolis is a unique source of legal status information which goes beyond the EPO’s INPADOC legal status content.
- Minesoft PatBase: Machine translated abstracts for both Japanese applications and utility models are produced for documents which do not have any other English language family member available (e.g. newly published documents); these are replaced with hand translated abstracts if/when they become available. PatBase also includes F-I and F-term [7] Japanese classification marks in its database, and according to system help documentation, this classification data is updated every two weeks. The PatBase Japanese collection includes both “clipped” (representative) drawings for each patent document, and embedded in-text images (such as chemical structure images, for example). Finally, it’s relevant to this discussion that PatBase’s WO/PCT collection includes Japanese-language Patent Cooperation Treaty (WO/PCT) [8] applications; this is rare, as most commercial vendors do not include PCT documents published in non-Latin character languages in their collections such as Chinese, Japanese, Russian, or Korean, even though they are designated official languages under the PCT.
- Thomson Reuters Thomson Innovation: Thomson Reuters’ machine-assisted translation collections are updated monthly. Representative images are included for the published applications collection back to 1980. Thomson Reuters added Japanese F-I and F-term classifications in a July 2009 update [9] to the system.
- LexisNexis TotalPatent: Full text English machine translations are not yet available for this collection; however, the TotalPatent coverage page lists machine translations present from 1975 for applications and 1996 for grants. This may refer to machine translations of bibliographic records created to supplement the EPO’s DOCDB file (also sometimes called the INPADOC [10] bibliographic file). The coverage tables for TotalPatent are generated by algorithm rather than by hand, and their earliest dates of coverage are often quite early – however, their date ranges do not indicate comprehensive coverage over the specified time period; a 1913 start date could indicate data for a single document published in 1913. This is true for all coverage tables, but is especially important to keep in mind with TotalPatent’s early start dates.
Thanks for bearing with me during this in-depth investigation into the Japanese patent collections offered by these major search providers. Although the material may seem a bit dry, understanding differences in search system coverage can be critical to performing a thorough prior art search. For further in-depth information on any of these systems, I hope you’ll stop by Intellogist.com to check out our detailed Search System Reports [11]. Comparisons of the major features of these systems can also be generated from our Quick Table Comparison feature [12] (free registration is required).
Finally, many thanks to Patent Baristas for hosting this series!
Today’s post is by Guest Barista Kristin Whitman [13] of Intellogist.com. [14]