Google Patent Search Hiccups?

Add to DeliciousAdd to DiggAdd to FaceBookAdd to Google BookmarkAdd to RedditAdd to StumbleUponAdd to TechnoratiAdd to Twitter

[tweetmeme source=”Intellogist” only_single=false]

Recently we’ve been noticing that Google Patent Search has been giving us some odd errors.  Specifically, some of its patent PDF copies seem to be only sporadically available.

To try to replicate the problem, we’ve been throwing some random patent numbers into the system, and while we haven’t figured out exactly what the database is missing, we have come up with some examples.  This past week when viewing the record for US7524860 (issued April 2009), clicking “download PDF” produced the following error message:

Google Patents Error Message

US7524860 error message, generated August 11th 2010

Another test document US7576097 (issued August 2009) produced the same error last week.  We contacted Google Patents via their contact form, and while we haven’t heard back, the errors (with these two examples, at least) have been corrected.

Of course, the good news is that there are plenty of other free sources for US patent PDFs should one temporarily have issues with one of them, including FreePatentsOnline, SumoBrain, Patent Lens and esp@cenet for example.   Has anyone else noticed this trend with Google Patents or any other system?

Like This!

Technical Translations from Landon IP

This post was contributed by Intellogist team member Kristin Whitman.

8 Responses

  1. http://www.ipr-helpdesk.org/newsletter/31/html/EN/howToFindAPatent.html

    Adds some relevant remarks. Some issues have been solved since the publication of this review, while others remain valid.

  2. Thanks for the helpful link, Paul!

  3. Well, there goes my only reason for recommending Google Patents. Unfortunately, it appears the USPTO is comfortable with Google’s uneven quality and spotty service.

  4. I still find it to be a great kickstart to any patent search (or a second take when stuck finding a reference) because of the “Google Magic” relevance ranking that brings good references to the forefront.

    I definitely would not count on it to be my only source.

  5. Please, PLEASE, will everybody from now on distinguish between “free” and “free” sources for patent information.
    Firstly, there are the sources which are in some way part of a commercial enterprise, and get income from mouse clicks, advertisements etcetera.
    Secondly, there are the sources which give patent information because of the classical legal trade-off: “you get the patent rights if you publish the information through us”.
    This second source is of course the collection of national and supranational databases: USPTO, esp@cenet, WIPO etcetera. For a limited list of Internet sources see my company’s website.
    I always recommend any non-professional searcher to use only the second group, and to be very reluctant with the first. For obtaining pdf’s I recommend http://www.poxoq.com, where you can download a versatile piece of software for quick downloading of patent documents.

  6. You make a very important point here. Although it’s obvious to information professionals, it may not be obvious to casual searchers that you must *always* scrutinize the agenda of any website that makes information available for free.

    It’s absolutely true that free patent search sites such as FreePatentsOnline (http://www.intellogist.com/wiki/Report:FreePatentsOnline), WikiPatents, Patentstorm, use SEO techniques to generate lots of hits using full text patent collections. They do exist to make money and they don’t necessarily have the interests of the public searcher at heart. When you consider the economic motive, it seems very possible that the reliability of the information they offer can suffer because of that.

    They’ll never reach the level of customer-orientation shown by the national and regional patent office services that you mention. However, it’s still a good idea to figure out how the patent search community can benefit from the services they offer.

    Google Patents (http://www.intellogist.com/wiki/Report:Google_Patent_Search) is kind of an interesting case because so many of the large-scale information projects Google has put together have greatly benefited the public. However it’s still an important point that they are a company with a profit motive. I believe they offer value above a FreePatentsOnline or PatentStorm type of service because they provide such fast load times for the patent images – although they have major limitations in timeliness and of course they don’t offer the current US patent classifications.

    In addition, some free services exist to provide playgrounds for testing new search algorithms. I’m thinking of PatentSurf (http://www.intellogist.com/wiki/PatentSurf) and Xyggy Patent (http://www.intellogist.com/wiki/Xyggy_Patent) here, but I’m sure there are others. Because they exist to showcase technology it’s important for these services to work very well, so this type of free product can benefit the community also. I would never rely on one of these services exclusively, but they can provide interesting supplements to an existing search – it’s reasonable to assume that they may be able to turn up wildcards or unusual hits.

  7. To get back to the original error message (Not Found), at least it wasn’t the Dread Pirate 404 (Page No Longer Exists) error! That’s a solid data result, and not customer service muzak– I’ve noticed recently that where you formerly were just placed in limbo at USPTO PAIR during high volume, now you usually get notified they can’t do it and to try again shortly– and it’s only a matter of seconds before you can connect. However, I’ve also received a message there and elsewhere that there’s no such thing (the dreaded “No Results,” when actually a search wasn’t run at all). When I know that’s a crock, I try again, and often get right through.
    I haven’t been able to piece together at what level this is happening– seems to happen ab initio with different search engines, at different sites, at fluctuating times; but I do think saying “I don’t know” when you’re too busy or important to be bothered may be a wrinkle to contend with. Zero or none has to be defined, to know whether a query was executed, or whether you never got through the gate. That Google message said it wasn’t found “on this server”– and all services admit that when they’re uploading or doing maintenance, they can’t walk and chew gum at the same time. A better message would let you know what’s happening– like the cute Twitter bird carrying the whale, “Too many tweets!”
    I’d bet good money on this scenario, Google tries to reach its imaged pdf cache, and when timed out (Google’s so ADHD!), says it’s not available. Maybe somebody changed the timing algorithm, knowing that if the server doesn’t answer in two seconds, it’s not going to answer in twenty– and doesn’t care if it’s because a tsunami hit or because somebody didn’t push to On button that day. Do let them know (quick Contact Us message) when you hit problems, or they won’t improve.
    The take-home: these are real-time services, not static libraries, and rely on many interactive connections. Not Found happens. My routine is not to believe them, and to try one more time– reminds me of the mistimed RFID in my TV remote, which turns on the monitor before the set-top box fully activates, which in turn turns off the monitor which turns off set-top box. A lot of flashing lights with…yep, No Results. (And don’t get me going on patent thumbnails…)

  8. Kris, I completely agree we need more transparent error messages! Why is this so hard to accomplish? I suppose the ideal situation is to believe your users will never see an error message, but you’ve got to plan for the worst.

    I laughed when you described the search engines as “too busy and important to be bothered.” Have you ever seen Bridget Jones’ Diary? There is a great quote from that movie – “Shut up please, I’m very busy and important!” Google patents should make that one of their go-to error messages, at least we’d get some entertainment value out of it.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: