KMWorld CRM Media Streaming Media Faulkner Speech Technology Unisphere/DBTA
Other ITI Websites
American Library Directory Boardwalk Empire Database Trends and Applications DestinationCRM EContentMag Faulkner Information Services Fulltext Sources Online InfoToday Europe Internet@Schools Intranets Today KMWorld Library Resource Literary Market Place OnlineVideo.net Plexus Publishing Smart Customer Service Speech Technology Streaming Media Streaming Media Europe Streaming Media Producer Unisphere Research



News & Events > NewsBreaks
Back Index Forward
Twitter RSS Feed
 



Google Tackles Patent Searching
by
Posted On January 2, 2007
Like Alexander looking for new realms to invade and conquer, Google has entered the IP (intellectual property) arena, specifically the subset of patents. The new weapon (excuse me, product) is Google Patent Search (GPS), a beta service that provides access (for searching and for display) of more than 7 million U.S. patents. According to the FAQ, patents from 1790 to mid-2006 are included, but U.S. patent applications are not. Patent documents were OCR'd into the database; the usual caveats about this technology apply. Google said that it plans to expand the content to patent applications, international patents, etc.

In the commonwealth of information, patents are the mysterious realm—at least to the public and to much of academia—rife with both jargon and argot (remember, they're different—jargon is a matter of consistent definition; argot, in this case "legalese" or "patentese," may be intended to be obscure). Dangerous territory for the tyro or uninitiated, but GPS (pardon the multimeaning abbreviation) may be a good weapon to tackle the dragon. On the other hand, the IP "white knights" (read professional patent searchers) may see fit not to use this resource. However, wise knights would do well to evaluate the service so that they can respond to questions from their customers or from the general public, such as "Just how good is this?" or "Have I found everything I need to know?" In fact, the real danger is that nonprofessional users of GPS will believe they've searched everything there is.

One group not yet heard from is the open access crowd. Hopefully they will not just hail this as another free resource for the masses but will also consider problems with accuracy and venues for user input.

J. Q. Public is most likely to stumble into Google Patents by spotting the "Patents" button under "more" on the Google toolbar (http://google.com). Simple keyword searches can be performed just like on any other Google product. However, better results can be obtained on the advanced search page (from the home page or at www.google.com/advanced_patent_search). Text searching can be done with implied and/or/phrase/not logic, plus these items can be searched separately or in combination: patent number, title, inventor, assignee, current U.S. Classification, International Classification, and issue or filing date ranging. Hit-term highlighting, albeit pale yellow, is shown on the image file.

I was first alerted to this new service by sharp-eyed patent searchers in the chemical-information and patent-information communities on the CHMINF-L and PIUG (Patent Information Users Group) lists. Interest was vigorous and critical. Several errors were easily found—many were due to OCR errors, others were not so easily explained.

Then I was offered the opportunity to write a NewsBreak. Thus intrigued, I gave GPS a try myself. Rather than search a "biggie" such as Xerox or Exxon and check for retrieval numbers, I settled on a corpus that I was more familiar with: my three patents (vestiges of my career in the laboratory; I should have more like 20). Searching "Robert Buntrock" as inventor retrieved two of the three: US 3784555 and US 4876044. The first has a strange title: "CERTAIN OXADIAZOLYL AND TFFLADIAZOLYL AMDENES." I realize that even the correct title may seem strange but it is "Certain oxadiazolyl and thiadiazolyl amidines." Misspellings both gross and minor occur in the last two words. Correct spellings occur elsewhere in the document, so a text search would retrieve the patent.

The missing patent, US 3874874, can be found by searching the number; examination of the image gives no clue why it was not retrieved. A remote possibility is that this patent is a division of US 3784555, and such documents were not included. To date, Google has not responded with a reason.

Other strange errors can be found. The reporting format is basically good, but results of a search for an inventor appear as "ininventor:Buntrock."

Examination of the image of US 3784555 also gives no clue why the misspelling occurred. The phrase "THIADIAZOLYL AMIDINES" is spelled correctly in the record for US 3874874, so that tongue twister is not the reason. C'est la guerre .

Several comments and gripes have already been reported by subscribers to the lists cited above (archives are accessible at http://listserv.indiana.edu/archives/chminf-l.html and at http://piug.org/list.html). Retrieval counts for various searches (such as keyword, inventor, assignee, etc.) show that GPS often underperforms other compendia of U.S. patents. OCR errors are fairly common in GPS, which can be quite damaging to retrieval even if they are rare on a percentage basis.

Professional searcher Edlyn Simmons from Proctor & Gamble reported searching "Google" as the keyword. Retrieval included an apparent inventor named Google (1927), a spelling error ("SWIM GOOGLE …"), and a patent with a nonsense title (X's) assigned to Yahoo! that cites references retrieved by a Google search (?!).

Carol Sidey from the Xerox Research Centre of Canada reported an apparent limit of 1,200 hits when other sources will retrieve far more (e.g., "Xerox" on GPS turned up 1,204 hits, while the USPTO file listed 17,390).

Donna Kleiner from Symyx Technologies reported an interesting downside of the advent of GPS on another Google service: She began to retrieve patents on Google Scholar even when they were not wanted. Currently, there seems to be no mechanism for eliminating them from the retrieval.

Roy Zimmermann from Medtronic reported on both favorable and unfavorable aspects. Along with several others, he described GPS as a good "quick and dirty" resource, particularly for pre-1970s patents. However, like most professional responders, he believes that GPS will never be a primary resource for professional patent searchers.

Several of these professionals wondered about the impact of GPS on their clientele—if it would represent "Wal-Marting" the industry and siphoning off users with Q&D (quick and dirty), free but inadequate search results. However, Jim Johnson from The Chamberlain Group pointed out that the "Wal-Mart effect" isn't always bad for existing business. He noted that the entrance of Wal-Mart and other "big-box" stores in several small towns impacted local businesses negatively to some extent, but in the long run, the overall business climate improved.

However, information industry analyst Bette Brunelle of Outsell, Inc., thinks that Google Patents will not succeed because it won't penetrate the relatively small patent-information market. She also indicated that the fate of beta versions of other free Google searching services was not favorable for improvement.

How receptive is Google to input? The Google Patents FAQ suggests emailing suggestions for improvements to patent-support@google.com. Some users have reported favorable responses to their notes on errors and omissions. However, there's no track record yet on whether or not the errors can or will be corrected.

So, will Google Patents continue to impact like a Wal-Mart, or will it just putter along or even fizzle? Only time (and apparently, the broader user public) will tell.


Bob Buntrock is a (semi)retired chemist and informationspecialist living in glorious Maine. A veteran of the online revolution and wars, he reviews books and writes on a number of topics.


Comments Add A Comment

              Back to top