Yeah, in the last 16 years we’ve gone from the classification leader to the classification follower… so too in other patent areas as well.

Yes, G9B comes in all sorts of handy. Thank you for showing me that way back when!

And I agree with regard to classification. Too bad that has been contracted out to semi-trained monkeys. On the other hand, the way practioners draft claim 1 these days, anybody would have trouble classifying them. Funny thing is that Europe used to look to the constant US reclassifications as one of our strong points back in the day.

I checked the earliest 20 PUPUBS from 2001 and the most recent 20 PGPUBS from 2005 with the g11b5/31.cipn. designation: 70% of the earliest 20 and 80% of the most recent 20 appear to be clearly incorrect classifications, though perhaps the contractor would call them “liberal” classifications.

I think restoring some order and integrity (and forward-looking vision) to patent classification should be a first order of business for Mr. Kappos et al.

Re: ipcr. I think the ipcr command will capture the PG-Pubs that have been reclassified during prosecution while the cips and cipp will only capture the intially implemented classifications. Since I’m not sitting at my PSR terminal testing this theory on a Sunday (thank goodness), I may be wrong… But it does capture additional references over the other commands [ Enter as g11b5/31.ipcr. or g11b005/31.epc,ipc. ] and is easier to understand by a non-searcher reviewer (read: attorney client) since it has the letters ipc in it.

Re: multi-generational F/B searching combined with a date limitation. Yes, a great way to start any validity search! Now don’t give away all out tricks and techniques. [ And don't forget to include the various application date formats in your cutoff ]

Re: family grouping. If I recall correctly, this is one of Questel’s significant positives over their competition. But as I said, I’m an EAST user, so I do not remember all the various product differences.

That said, we both know every patent searching engine has pluses and minuses. And if you’re a good searcher, you’ll find a way to take advantage of each engines strengths. I’ve been to enough demos and tried enough of ‘em to believe this! Now, hopefully this comment doesn’t spend 2 days in Gene’s moderation box.

DIP

Questel-Orbit introduced a new dimension of forward/backward citation searches around 2000 (it goes well beyond EAST’s 1-dimensional US capabity). Perhaps the EAST developers might use it as a model. In Questel’s Pluspat, you can easily 0) obtain all the family members for a document and then (for all US and foreign family members):

1) perform backward citation searches that includes all foreign documents,

2) perform forward citation searches on all the backward citations found in 1,

3) perform forward citation searches that includes all foreign documents,

4) perform backward citation searches on all the forward citations found in 3,

5) AND all the results together and then limit them by publication or application date.

It takes about 2 minutes (there can be MEM/print charges in Pluspat, so below may not be efficient). In one fell swoop, you learn all the art many applicants thought was best (and were required to disclose under Rule 56), as well as the best art that many Examiners knew about from their individual experiences. What a way to start a search, without even using keywords, to learn what’s been known across the industry (and perhaps across the world) before!

[This is not legal advice:]

file pluspat
/pn us7123456 or ep1234567
citb ss 1
mem /pn rk 4
prt mem mem1 nonstop
er mem mem1
citf ss 2
mem /pn rk 4
prt mem mem1 nonstop
er mem mem1
citf ss 1
mem /pn rk 4
prt mem mem1 nonstop
er mem mem1
citb ss 4
mem /pn rk 4
prt mem mem1 nonstop
er mem mem1
1 or 2 or 3 or 4 or 5
6 not /pd>2001-02-03
mem /pn rk 4
prt mem mem1 nonstop
his

Regarding “grouping” of family members, I would never suggest (or consider) viewing just one family member for an infringement search (as EAST currently provides with the WPI file). Rather, the record of EVERY single family member should be concatenated into one family record, so (for example) all the claims of all the family members are immediately at your disposal when you say, “Display the claims.” For example, in the family record, first you would view the US patent (if it exists), then (if they exist), the US published application, the US related patents and their published applications, and the counterpart foreign patents (JP B, EP B, etc.) and published applications (WO, GB, etc.). When text searching, for example, the indexed full text of the PGPUB and the foreign pubs/patents (with machine translations) would be concatenated to the full text of the US patent(s).

If the text from ANY of the family members satisfies a text search (or if the classification satisfies a classification search, etc.), the whole family text record would be displayed, with the appropriate KWIC highlighted wherever it appears. If you don’t want to see the foreign equivalents or foreign information, just don’t go there – flip to the next document – you’ll be done with that family. (The concatenation order in the family record could be adaptive, based on which documents caused the family to be considered a “hit.”)

Good post re: EAST. I’m a full time user of the system and it is great for being free and super fast. But it also has very distinct weaknesses. US Examiners and searchers using EAST have questionable access to foreign art. We may have abstracts of 30 million documents, but the multiple databases include a ton of overlap as you mentioned, no updating of the EP database (replaced by WPI perhaps), no full text searching, and virtually no old (pre 1980) foreign art.

That said, it is an old system and hasn’t had a major update in 10 years. But at least they’re finally letting us avoid reviewing PG-Pubs that have a corresponding issued patent after eight years of our complaining! And the ability to perform large Forward and Backward citation searches as a separate L# is a powerful tool not available many places.

One thing to watch out for though if they do a major overhaul of the IT. Many of the very good 3rd party systems that simplify life by grouping family members for easier search purposes are doing you/us a disservice for clearance/infringement work because the claims of each of those family members are unique and distinct from one another. And while I love semantic analysis and relevancy rankings as yet another search tool, it’s very easy to fall into the trap of using them exclusively, much the way many examiners/searchers do with key word searching.

On another note Scrappy: Use .ipcr. rather than the cipp and cips codes for pulling art by IPC

Senator Franken asked him about copyright issues. I was the one who said it was stupid for him to be at all concerned about copyright issues. I stand by that comment. There are plenty of other people in the US government who are in the entertainment industry pocket. Obama has appointed several of them to top DOJ positions, there is the Copyright Office and Trade Representative who can also handle copyright issues. None of them are at all qualified with respect to patents. Is is ridiculous for the Director of the Patent and Trademark Office to also have responsibility for copyrights, regardless of the title he holds.

So if the Senate or President expect Kappos to utter a single word about copyrights then they need to be ready for the consequences. I will berate them daily and show to everyone just how unimportant the US economy and innovation must be if the top innovation official in the government has to pander to the copyright industry and divert attention to curbing piracy when so many others in government already have that as their primary job description.

-Gene