Ready or Not, the Cooperative Patent Classification Has Arrived!

[tweetmeme source=”Intellogist” only_single=false]

The patent world is about to get rocked by the upcoming 2013 roll-out of the new Cooperative Patent Classification (CPC), which will replace both the current US patent classification system and the European patent classification system (ECLA).  This rollout is coming so quickly that after January 1st, 2013, the CPC will completely replace ECLA, and all US published applications (also known as PG-Pubs, or A documents) will carry CPC classifications.

The US Patent and Trademark Office held an External User Day Event on July 10th at the US Patent and Trademark Office in Alexandria, VA.  This event was an opportunity for the USPTO to listen to input and concerns from the patent search community regarding the CPC.  Upon invitation from the USPTO, Landon IP’s Director of Training and Special Projects, Jonathan Skovholt, served on the External User Panel to provide his analysis and comments on the effects of the planned transition.

How will these changes affect you? Read on to learn about the features of the new system, and Jonathan Skovholt’s analysis of its possible effects.

More valuable information was presented at the External User Day program that can be covered in this post, but I’ve chosen some valuable highlights about the features, advantages, drawbacks, and effects of the CPC implementation.


Features of the CPC

Extension of the IPC System: The CPC has been developed as an extension of the International Patent Classification (IPC), which is often lacking in detail and contains broad classes overstuffed with patent documents.  Although other IPC extension classifications exist, the CPC surpasses all of these systems in granularity, with over 260,000 individual codes, as compared to:

  • 69,000 for the International Patent Classification (IPC)
  • 145,000 for the European Patent Classification System (ECLA)
  • 150,000 for the US Patent Classification System (USPC)
  • 180,000 for Japanese File Index (FI)

(Note that the CPC still doesn’t come close to the Japanese F-term class system, which has over 350,000 subdividsions.)

Detailed Class Titles: Another targeted improvement is that the CPC subgroup titles are typically more descriptive than the USPC titles, aiding comprehension for those who scan the codes looking for relevant materials.


Analysis from Landon IP’s Jonathan Skovholt:

Although the idea behind the CPC was to develop a classification system with a high level of granularity, Landon IP’s Jonathan Skovholt noted that the CPC hierarchy loses one major advantage of the ECLA system:  ECLA’s subcodes got longer as their level of granularity increased, and followed a letter-number-letter-number organization.  For example, see the following hierarchy, which gets more detailed as the codes increase in length:

H01L 21/285

H01L 21/285B

H01L 21/285B4

H01L 21/285B4A

This structure meant that it was easy to gauge the specificity of ECLA codes at a glance, especially when you were looking at classes out of context.  CPC, on the other hand, flattens out the hierarchy and drops the easy visual cue of the letter-number format.  See the following example, which illustrates a sample transition from increased length to fixed length codes:



H01L 21/285 H01L 21/285
H01L 21/285B H01L 21/28504
H01L 21/285B4 H01L 21/28508
H01L 21/285B H01L 21/28512

The loss of this visual distinction between codes is the loss of an ECLA feature that some found helpful in determining the relative granularity of the code at a glance.

A second issue is that the professed goal of the CPC project is to eventually have other patent offices adopt the system.  Mr. Skovholt pointed out that this goal may require further targeted classification activity, because patenting in other countries can be heavily weighted towards specific technologies that may not be as well covered in the US and European areas.  For example, the Japanese art can be said to be extremely heavily biased towards the automotive field.  The revision committee will have to put special emphasis onto discovering and classifying these technologies thoroughly if they expect global adoption.


Reasons for CPC Development and Requirements for Implementation

The US and European patent offices were both interested in developing a joint classification system to reduce examiner workload.  Harmonization is expected to:

  • Enhance efficiency by requiring Examiners to use one system, rather than having to use both systems for maximum retrieval
  • Improve navigation and understanding of patent classifications by removing the complexity introduced by multiple classification systems
  • Facilitate work sharing on patent applications filed in multiple IP offices
  • Lower the costs of classification work
  • Improve the consistency of classified search results across IP offices
  • Provide an adaptive and actively maintained classification scheme

Implementing the CPC will require both offices to invest time and money.  Both offices must:

  • Develop new training programs to ensure adoption and correct use of the system
  • Collaborate on CPC revisions
  • Update IT systems to support CPC searching

Analysis from Landon IP’s Jonathan Skovholt:

It’s not a given that patent search efficiency will increase with the introduction of the CPC.  In the short term, professional patent searchers will see an increased classification burden, as a thorough search will require searching all existing class systems as usual.  Based on the USPTO’s timeline, it will be quite a while before the full transition over to CPC has ended: because only US published applications will be classified from Jan 2013 onwards, it could be as late as 2016 before all issued grants bear CPC classifications.  In our experience, old tools retain their utility even after new tools are introduced, and the end result is often that the toolbox just gets bigger, not smaller.

Even after the transition, there is the possibility that other classifications are found to have advantages over the CPC in certain situations, and the static/legacy class systems will still be used to supplement patent searches.

Although the goal of providing an adaptive and actively maintained classification scheme is admirable, it’s worth noting that collaborative revisions must both be approved by both the USPTO and the EPO.  This approval process could prove to be a bottleneck if either of the offices can’t make these revisions a top priority.


Re-classification of the Existing Patent Corpus

Because the CPC is a new classification system, and not just a copy of ECLA with some USPC codes added, there will not be an exact 1:1 concordance or auto-conversion engine that can take an old code and show the new CPC equivalent, or vice-versa.  Users should not expect a concordance engine to be provided by either office.

However, the conversion process of already-published documents is expected to proceed quickly. Many US documents are already classified in ECLA, and because the CPC and ECLA have common roots, this should make the mapping process relatively simple for both US and EP documents.

A larger problem will lie in the re-classification of the older US backfile.  US re-classification efforts currently extend back to documents published in the 1920s.  According to the USPTO representatives, the re-classification process is mostly automated.  Mapping software will perform analysis of the document’s existing classification codes, and will also perform a keyword analysis.  In the case of a one-to-many or many-to-one relationship between codes, some intellectual re-assignment could take place by hand.


Analysis from Landon IP’s Jonathan Skovholt:

The introduction of a new patent classification system is always a scary prospect when you consider that millions of existing documents have to be re-classified.  The success of the CPC depends heavily on the quality of these re-classification efforts.  If a significant percentage of documents are misclassified, searchers will have to heavily rely on the older systems to perform competent, thorough searches in the older art, making the introduction of the new system far more of a burden than a blessing.


Additional Comments from Kristin Whitman:

One comment made at the External User Day panel came from a USPTO official who suggested that if poor re-classifications took place, the examiners who search heavily in the art unit in question would re-assign the patents as necessary.  To me, the idea that heavily-used art will be more likely to be properly classified is not a sufficient way to address re-classification concerns.  It’s not only heavily cited art that is valuable to an in-depth project such as a validity or freedom-to-operate search.


New Practices at the USPTO

Examiners at the USPTO will have to be flexible when adapting to the CPC.  Although a full training regimen is in development for Examiners, the shifts in practice may take a while to fully come together.

The CPC will change certain elements of USPTO practice.  First, Examiners will need to be re-trained on classification methodology, since the CPC requires classification of all inventive material, not just claimed material.  In addition, the CPC could affect how US Examiners are able to issue restrictions, as they must be taken into consideration when proving that a restriction is necessary.



By reaching out to external users, both offices are showing a commitment to making the CPC a success.  The External User Day provided an open forum for searchers to submit questions and concerns.  The day’s full presentations will soon be available on the official CPC website.

What are your concerns about the CPC transition, and how would you respond to Jonathan Skovholt’s analysis?  Leave your thoughts in the comments section!

Technical Translations from Landon IP

This post was contributed by Landon IP Librarian Kristin Whitman. The Intellogist blog is provided for free by Intellogist’s parent company, Landon IP, a major provider of patent search, technical translation, and information services.

16 Responses

  1. not to pound too heavily on us examiners, but classification into ipc and presumably ecla has historically been poor. i acknowledge that if i had to classify into uspc it would be poor too.
    i just hope that reclassification is manual and not automated or done with a concordance tool. it’s just not accurate that way.
    i don’t think less detailed sets of classes relating to a subject matter will deter too many. the same situation exists in the ipc today. for example the c07d400 series has barely changed in 40 years and they are horribly broad.

  2. ECLA is an excellent classification system much superior to that of the IPC. It is probably the best that is available.

  3. just hope that reclassification is manual and not automated or done with a concordance tool. it’s just not accurate that way.

  4. I was an examiner for over 35 years and when I was told about the conversion, I was curious about the accuracy of the conversion of the current US patents, especially the older ones. I was told by the officials at the PTO the EU has done a “wonderful job”. I picked a subclass and checked the older ones and how they were classified. The area I picked was 244/51 AIRCRAFT, STEERING PROPULSION: Devices and arrangements for steering aircraft in flight by means of the propelling agency. The older patents showed a lot of airships with nothing novel about the airship, but a particular steerable propulsion that was properly in the USPC system at 244/51. I went to the espacnet site and confirmed what the new ECLA classification was. The “great job” done by whatever automated or EU classifier was not very good. Almost all were classified in an area unrelated to steerable propulsion and many in the broadest area of airships B64B 1/00. Most of these patents showed an airship in figure 1 and the steerable propulsion in later figures. I checked some other areas and found more correctly classified patents. However as I checked many more I started to see a pretty good correlation to a classification of the older documents to what was seen in the first figure, if it was a good depiction of the invention it was correct if it was a more general depiction without specifics it was classified incorrectly to what was see generally in the first figure. I did not see many multiple classification in older documents even though later figures showed more inventive concepts. The newer patents seemed to have a lot better accuracy, I think the older ones did not get a lot of attention to detail. Anyone interested should check for him/her self in an area they are familiar with; I think the results will be repeated with some substantiation.

  5. aero man
    we both seem to agree that reclassifying by concordance in both directions isn’t always the best, so it seems to reinforce the ‘best practice’ of searching the new classes as well as the old, supposedly reclassified, classes.
    for example i’d search 244/51 and b64b1/whatever, as well as whatever the cpc class is (as long as the database you’re searching holds historical classifications)

  6. […] created by the EPO, such as the free online patent search system Espacenet and the upcoming Cooperative Patent Classification system (CPC).  The EPO  recently announced updates to both of these products, so I’ll give you a brief […]

  7. […] (ECLA), which has already been been frozen as of July 2012 and will soon be replaced by the  Cooperative Patent Classification System (CPC).  The full rollout for the CPC is January 1, 2013, but the EPO and USPTO have already released the […]

  8. […] to the IPC classification system, which was frozen in July 2012 and will soon be replaced by the Cooperative Patent Classification (a joint effort between the USPTO and […]

  9. […] year for new patent search products and updates, and 2013 updates will be even bigger due to the release of the CPC.  See you in 2013, and have a Happy New […]

  10. […] Ready or Not, the Cooperative Patent Classification Has Arrived! – Kristin Whitman gives a detailed overview of the of the features of the Cooperative Patent […]

  11. […] been following the development and roll-out of the Cooperative Patent Classification (CPC) over the past year, from the initial release of the CPC scheme to a guide describing how major […]

  12. […] ranging from the addition of a Japanese patent full-text database to the implementation of the new Cooperative Patent Classification (CPC) within the portal. STN also recently announced the beta release of the first version of a new STN […]

  13. […] January 1, 2013, the Cooperative Patent Classification (CPC) completely replaced ECLA, and all US published applications will now carry CPC classifications. The […]

  14. […] AP documents, the OA documents (even without family members) seem to have been updated with the new Cooperative Patent Classification (CPC) codes.  And on the plus side, it does seem that original document PDF copies are available for […]

  15. […] some of the AP documents, even those without family members, seem to have been updated with the new Cooperative Patent Classification (CPC) codes, perhaps suggesting that someone (I assume the EPO) had been categorizing these documents […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: