OT: Communicating with Google about errors in the Google Books archive

Garson O'Toole adsgarsonotoole at GMAIL.COM
Mon Mar 15 19:58:22 UTC 2010


Joel S. Berson
> Thanks, Garson -- I might try this sometime.  Apparently not all
> feedback with corrections ends up in that big waste bin in the
> ether.

The response I received was not about metadata or page scan errors. It
concerned a question about the Google Books interface and software,
and I was glad to get a thoughtful human response. My reports about
poor page scans and requests to unprotect public domain (based on
date) works receive automated responses.

Victor Steinbok wrote
> Although Garson has mentioned this once before, there is a general
> problem with this approach--there are simply too many errors to report
> them individually. When in the process of single search--especially
> for a word that is either common or commonly misinterpreted by
> OCR--there can be dozens of errors just within a set of research
> alone, let alone the products of multiple searches. If Google really
> has any interest in fixing problems, it should have a check-off list
> accompanying every entry--e.g., have check-boxes for "incorrect date",
> "incorrect title/author" (or other bibliographic info), "multiple
> books scanned", "missing pages", "mis-scanned pages" (or other
> scanning errors), "incomplete bibiliographic information" (i.e.,
> impossible to determine the actual book), "incorrect classification"
> (limited preview rather than full scan or vice versa), "incorrect
> subject categorization" (the error rate for this one is over 80%
> easily), "OCR errors". I am far less concerned with OCR errors than
> the rest of the menagerie.

Easier reporting mechanisms are desirable. Yet, the task of informing
Google about metadata problems is almost Sisyphean. Geoffrey Nunberg
covered the topic of poor-quality information in a Language Log post
and a Chronicle Review article several months ago.

Google Books: A Metadata Train Wreck - August 29, 2009
http://languagelog.ldc.upenn.edu/nll/?p=1701

Google's Book Search: A Disaster for Scholars by Geoffrey Nunberg -
August 31, 2009
http://chronicle.com/article/Googles-Book-Search-A/48245/

Jon Orwant, manager of the Google Books metadata team, responded in a
comment on the Language Log blog.
http://languagelog.ldc.upenn.edu/nll/?p=1701#comment-41758

Garson

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org



More information about the Ads-l mailing list