Discussion of all things patent mapping and analytics.

  • Home
    Home This is where you can find all the blog posts throughout the site.
  • Categories
    Categories Displays a list of categories from this blog.
  • Tags
    Tags Displays a list of tags that have been used in the blog.
  • Login
    Login Login form

Examiner vs applicant vs X vs Y citations - and does it matter? (Nov 2016 update)

Posted by on in Amberblog
  • Font size: Larger Smaller
  • Hits: 6142
  • Subscribe to this entry
  • Print


A comparison of applicant, examiner, X and Y citations for a Nike patent has shown that: 

  • The prior art comprised both applicant and examiner prior art
  • The examiner only appears to have cited the patents that were found by the PCT patent examiner
  • The prior art submitted by Nike contained, not surprisingly, some relevant prior art
  • Family Cluster Searching allowed us to instantly find some highly relevant prior that was not cited by the examiner
  • As confirmation of the similarity filter, it also confirmed the importance of the '102' citations nominated by the examiner. In this case, this suggests that the examiner was not wrong in his analysis - only incomplete.

Note - this blog was originally published in May 2014 - but has been updated to take into account our next generation Family Cluster Searching software




Introduction and background

Ambercite has built tools to find similar and important tools based on citation analysis. When discussing these tools, one of the questions that come up from time to time is: 

Can your system distinguish between applicant and examiner citations? Between X and Y citations?

This is an important question that is worth exploring. But first of all, an explanation of these concepts for those who are not fully across these terms.

Patent applications in the US generally come from two sources:

  • Patent examiners comprise a search based on the claims of the patent, and come up with a list of the prior art patents that they think are important. These are generally referred to as 'examiner citations'.
  • Patent applicants put forward a list of prior art patents that they know about, and which they they think are relevant ('applicant citations'). This is part of a Information Disclosure Statement (IDS). Sometimes these patent citations are found when the same patent being examined in other jurisdictions, and being notified to the patent applicant - who is then required by law to notify these to the USPTO. There can be more than one IDS for a patent being examined.

Within this group of citations, there is a subset of those patents which are specifically referred to by the examiner in their written examination reports. In the US these are sometimes referred to as '102' (relevant to the novelty of the patent, as per Section 102 of the US patent act) or '103' citations (relevant to whether the patent is obvious).

Outside of the US, for example in WO (PCT patents) or EP search reports, prior art patents are classified as 'X' and 'Y' citations respectively, where X documents relate to novelty and Y documents refers to inventive step.

Having divided up the available patent citations into these different types,  these is a natural temptation to regard some types of citations as being more important than others. For example, it has been reported that some patent examiners often ignore the citations put forward by applicants, instead preferring to rely on their own search reports. Certainly, the fact that an examiner has specifically referred to a patent in a written search report should carry some weight. 

But should this be to the exclusion of patents not specifically mentioned in a written search report?

Should patent citations put forward by patent applicants be automatically disregarded?

There would have been reasons for the listing of these patents in the IDS, admittedly some weak reason in some cases. But if we assume that patent applicants do the right thing and put forward all of the prior art patent they know about, some of these patents might be highly relevant. On the other hand, we would not expect this list to be complete - it is only the prior art they know about.

These are all good questions.

Another good question is - why does this matter?

In our case, Ambercite analyses patent citations not for the sake of it, but because we believe patent citations can help identify similar and important patents which can strengthen existing search processes. Also we try to provide new information not available by other means to our clients. In many cases our sophisticated clients already have access to information regarding the 'status' of citations - can the likes of our patent search tool AmberScope provide additional value beyond what is available within the classification of known citations into 'applicant', 'citation', '102', '103', 'X' and 'Y' categories, etc?


Case study on Nike patent 

To answer this question, we might consider the patent filed for the Google Maps technology, namely US7945343, for a Method of making an article of footwear, filed by Nike. 

A method of making an article of footwear is disclosed. The method includes the steps of designing an article of footwear via a website, converting a footwear representation into a set of two-dimensional portions, printing the two dimensional portions onto a sheet material, cutting and assembling the two dimensional portions into a finalized article of footwear. The method further includes a step of shipping the article of footwear to a pre-designated shipping address.

I have selected this patent because it should be a good test of this concept, namely it has: 

  • a list of applicant and examiner citations on its front cover
  • A PCT family member, also with a search report
  • An EP family member, also with a search report

The front page of this patent lists a number of citations, and with some of these citations being starred, indicating that they were cited by the examiner. These patents are listed below - and therefore the remaining citations must be 'applicant citations' by definition.

The examiner patent citations listed on the front cover are:

  • US5255352 (First listed inventor was Falk)
  • US6299962 (Davis) 
  • US2005071242 (Allen)
  • US2006111976 (Pompushko)

There was also a single non-final rejection published (7 August 2010) during the prosecution of this patent, which noted the the following patent as a 102 (novelty) citation:

  • US200507124 (Allen)

While the following patent as cited as a 103 rejection:

  • US2006111976 (Pompushko)

So what did the matching WO and EP patents show?

PCT patent  - WO2008077014

X citations (novelty)

  • US2005071242 (Allen)

Y citations (obviousness)

  • US6299986 (Davis)
  • US2006111976 (Pompushko)
  • US5255352 (Falk)

Along with a family member citation

  • US2006101671 (Berent)


The matching EP patent EP2094122 gave the same list of citations. It appears in this case that the US examiner has used the same patents in his list of 'cited by examiner" patents as were given in the PCT search report - i.e. they did not do their own search.


What can Family Cluster Searching find in the way of relevant prior art?

Family Cluster Searcing provides an instant means for looking for similar patents, both originally cited and also missed by patent examiners. It can be very easy to use - preparing a query is as simple  as entering the NIke patent number:



And the result?

The highest ranked patent found that was filed before the Nike patent was indeed the US2005071242 Allen patent, listed by the PCT examiner as the only X citation. No surprises there. 

The Davis patent US6299962 is listed in 5th position, the US5255352 patent is listed in 10th, while the tUS20060111976 patent was listed further down. Please remember though that none of these patents were listed as X citations

We also uncovered a good number of previously uncited but relevant patents - with this list headed by  WO2007024875, for a Method and system for providing customized footwear to a consumer, and then US20080010867, for Personalized footwear. We also found US5339252, for an Integrated system for foot measurement, last and footwear manufacture - filed back in 1994. This is not a new concept.



Tony Trippe, a highly regarded patent analyst, has looked at the role of applicant vs examiner citations and concluded that:

identifying the .. citations coming from the examiners specifically, provides an even higher likelihood of relevance.

Tony is probably right - examiner citations are probably more likely, on average, to be relevant. However hopefully what this study has shown is that applicant citations should not be dismissed out of hand if you are looking for relevant patents, particularly with a tool like Family Cluster Searching to quick find the most relevant of these citations. Can you afford to miss relevant patents in your search reports?



This analysis has shown that:

  • The prior art for a Nike patent comprised both applicant and examiner prior art
  • The examiner has only cited the patents that were found by the PCT patent examiner
  • The prior art submitted by Nike contained, not surprisingly, some relevant prior art
  • Using the similarity filter in AmberScope allowed us to quickly find some highly relevant prior that was not cited by the examiner (This would take less than 5 minutes with experience)
  • As confirmation of the similarity filter, it also confirmed the importance of the 102 citation nominated by the examiner. In this case, this suggests that the examiner was not wrong in his analysis - only incomplete.


In other words, if we had made the assumption that the only relevant prior art is that identified by the examiner as such (examiner citations) or that nominated as X and Y citations, what appears to be highly relevant prior art would have been missed.

Getting back to the opening question:

Can your system distinguish between applicant and examiner citations? Between X and Y citations?

Our response is - we think both are important to our clients. 

Rate this blog entry:

Mike Lloyd

Senior Vice-President, Business Development

Mike has a joint role of Senior Vice-President of Business Development for Ambercite, as well as acting as an IP Management consultant for Griffith Hack. Mike brings extensive knowledge and experience of IP management and more traditional forms of patent analysis to help apply NPA™ patent analysis for the benefit of clients. Mike has a background and advanced degrees in chemical engineering and IP law, and has studied and worked in New Zealand, Sweden, Finland and Australia. Mike has published in the areas of chemical engineering, patent management, the value of patents and the use of patent data in in a wide range of publications and forums, and presented on patent analysis in conferences in Australia, Singapore, the US and France.