Examiner citations vs applicant citations - which are better?

August 21 2017 Examiner citations are better than applicant citations right? 

Certainly, I have heard this 'rule of thumb' from many people over the years. But is this rule always right?

A recent US Federal Circuit case has provided a case study. As reported on IP Watchdog, the Federal Circuit invalidated that US7581688, filed by Whirlpool for a Blender with Crushed Ice Functionality was invalid due to prior art, even after the PTAB had ruled that it was not.


From our point of view, the most topical part of the case was the patent that was used to invalidate US7581688 in these proceedings - namely US6609821, filed by Sunbeam for a Blender base with food processor capabilities


Not that this surprises us here at Ambercite. If we enter US75581688 into our AI solution for finding similar patents, and run an validity search for the most similar prior art, as shown below:

US7581688 query.gif

US6609821 ends up at the top of this list of similar patents, with a similarity score of 18, as shown below:

So once again. Ambercite Ai has been shown to work. But was US6609821 an examiner citation?

Actually, no. 

If we look at the front page of this patent, we see that it has two examiner citations. US66909821 is listed as prior art, but not listed as an examiner citation - instead it was supplied as a citation by the applicant.

Front page.gif

So what is going on here? It has been reported by a team from Richmond, Stanford and Columbia universities that examiners appear to in many cases ignore the applicant citations. 

We can only speculate why, but one possible reason could be that sometimes there are so many applicant citations that the examiner thinks it is easier to start again from scratch.

And along similar lines, many patent searchers  confronted with a long list of applicant prior art, simply focus on the generally much shorter list of examiners citations.

But in this case, if had either the searcher or examiner had relied on examiner citations alone, they would have missed the prior art document that the Federal Circuit cited when invalidating the patent.

Which really challenges the  'Examiner citations are better than applicant citations' rule of thumb, doesn't it?

In contrast, Ambercite Ai offers another, and we suggest better solution to help prioritize the prior art. Its advanced AI algorithms can rank of the citations given to suggest which of all of the citations given are most similar - regardless of whether they are examiner or applicant citations, are most similar. 

All in a matter of seconds. 

And in this case (and many others), with pinpoint accuracy.

It is for this and similar reasons, that we believe that when it comes to Ambercite Ai,  'no search is complete without it'.

Interested in applying Ambercite Ai to your own patent searching? If so, please contact us for a personalised demonstration.

Mike Lloyd