[bksvol-discuss] Re: Requirements for acceptance -- the bottom line
- From: "Rui Cabral" <rui@xxxxxxxx>
- To: <bksvol-discuss@xxxxxxxxxxxxx>
- Date: Fri, 13 Aug 2004 19:49:55 -0400
Hi:
I would like to address a few points, (in no particular order)
Getting more volunteers would be ok, but it doesn't change why we're having
this conversation to begin with, (text quality)
(if you have a lot more volunteers who submit and vallidate inferior quality
books, then we are back to square 1)
Next, there is no reason even with older software that we can't achieve at
least, at least 99.0% accuracy on most books. Those days are upon us if we
try, we don't have to wait for "some day."
It comes down to technique and settings as much as the recognition engine
itself.
You do not necessarily need the latest version of openbook or k1000 to
achieve this. I have vallidated quality scans from people on this list who
I know are using 4 year-old software other then openbook or k1000.
I do not expect "letter perfect" scans.
On the other hand, when i have to guess at what someone is doing or where
they went, that is unexceptable. That is not a readable book!!
That is exactly what people like myself and others want to do, make these
readable, almost indistinguishable from a print copy.
Usually i will only vallidate stuff that's rated excellent.
I grabbed a book the other day (shelley it was yours) i performed all my
checks all the pages were there. I vallidated it in 20 minutes.
On the other hand, i have spent 3 or 4 hours on a book before as well.
Why you may ask. it had its problems but it wasn't bad enough to reject,
the problems could be fixed with a little TLC.
I really don't appreciate that our efforts get put down in order to appease
a few.
If that is bookshare.orgs "standards, I would like to know what the
mechanism is to change them.
Certainly, i have no issue with the other criteria, copyright info,
title/author, duplicate, proprietary copies,etc. Those are no-brainers.
Of course i'm put in a position where i'm supposed to reject a flawless scan
that has no author, but i can submit a book that is chock full of errors
(let's say 98%)
98% may sound great, but as some of us know that's actually pretty poor.
If it's an old book with a weird font and faded print and no one i mean no
one can scan it better then 97.5 or 98%, then fine by all means i will
escort that one in myself.
But let's face it, thats not the case 99 times out of 100.
Instead you get something like this, (this is the first book on my needs
rescanning list on the website)
"B is for Burglar by Sue Grafton could use to be rescanned. it's raided
"Fair" and is a bad fair."
Does it sound like I'm a little frustrated??? yeah, just a little.
-- Rui
rui@xxxxxxxx
----- Original Message -----
From: "Sarah Van Oosterwijck" <curiousentity@xxxxxxxxxxxxx>
To: <bksvol-discuss@xxxxxxxxxxxxx>
Sent: Friday, August 13, 2004 6:45 PM
Subject: [bksvol-discuss] Re: Requirements for acceptance -- the bottom line
> Great Idea. i have heard of so many sites that match volunteers with
> organizations that could use them, but of course i never remember them
after
> I finish reading the articles they are mentioned in. I should do some
> looking for those sites, and others could as well. I think a bookshare
> staff member might have to write up something or contact those sites to
get
> on their lists after that.
> I have to admit that sometimes I didn't want to explain to librarians what
I
> was doing because some people over react before they hear the hole story
and
> all the facts about the copyright law exemption, and also might feel that
I
> could damage their books, because they don't know me enough to know that I
> practically view library books as sacred. :-)
>
> Sarah Van Oosterwijck
> curious entity at earthlink dot net
>
>
> ----- Original Message -----
> From: "Richard Ely" <ely.r@xxxxxxxxxxx>
> To: <bksvol-discuss@xxxxxxxxxxxxx>
> Sent: Friday, August 13, 2004 5:03 PM
> Subject: [bksvol-discuss] Re: Requirements for acceptance -- the bottom
line
>
>
> > Ah,
> > There in lies the rub. That old #6. One readers notion of readable is
> another's of execrable. Much
> > has to do with two things, how badly do you want to read the text and
how
> many errors are you
> > willing to accept. It is wonderful seeing the care and concern that most
> on this list give to a text
> > they are submitting or approving. As a teacher with students who use
these
> books, I wish they were
> > all letter perfect. I think that such will be the case someday. For now,
> it seems we must accept a
> > less than ideal, but only three years ago, none of this awful problem
even
> existed. No books, no
> > problem!
> >
> > What does seem to be the case is the need for more volunteers to help in
> validating books. In the
> > days before electronic Braille production, there were thousands of
> volunteers across the country who
> > had passed their transcribers certification and who banged away at their
> Perkins braillers at home
> > cranking out books. Seems that there are many more computer literate
> retirees then ever. They do not
> > need to learn the Braille code to make books available. For the most
part,
> all they need is the
> > skills that many already have, proof reading and word processing. What
we
> need to do is to find
> > better ways to let people know what they might do in a few hours each
week
> at their own computers. I
> > am giving thought to donating the computer I am composing on to our
local
> library for the specific
> > task of providing a computer to volunteers in our little town to use to
> validate books. If Deb the
> > librarian agrees, I hope to start doing regular small group classes on
how
> to validate. What if
> > there were say 500 other libraries offering something similar. At the
rate
> of one book per day, five
> > days per week each library would validate 260 books. At capacity, such
an
> effort could validate
> > 130,000 books per year. That might go a long way in cleaning up that
> backlog of texts awaiting
> > valedation.
> >
> > Rather than picking the nits of weather or not a submitted book is good
> enough to accept, let's work
> > at increasing the numbers of validators, and the access points they need
> to do validation.
> > Rick
> > ----- Original Message -----
> > From: "Jesse Fahnestock" <Jesse.F@xxxxxxxxxxxx>
> > To: <bksvol-discuss@xxxxxxxxxxxxx>
> > Sent: Friday, August 13, 2004 5:10 PM
> > Subject: [bksvol-discuss] Requirements for acceptance -- the bottom line
> >
> >
> > > Hi everyone -- I have recently received many emails from volunteers
> confused about whether they
> > must reject a book or not. I think the higher standards maintained by
many
> members in this group
> > have confused people. Therefore, I will restate, as I did on Monday,
what
> a book must have to be
> > accepted:
> > >
> > > 1. The book is not already on Bookshare.org, or if it is, that it is
> being submitted as a superior
> > replacement or transcribed Braille copy.
> > > 2. The book is not an eBook acquired under proprietary agreement (e.g.
a
> commercial eBook, a book
> > from WebBraille, etc.)
> > > 3. The copyright name and date are included.
> > > 4. The title and author are included somewhere in the book.
> > > 5. The book is not missing multiple pages of core content (core
content
> does not include tables of
> > contents, indeces, picture pages, or other front or back matter and the
> like).
> > > 6. The book is readable.
> > >
> > > If a book meets all of these requirements, you may absolutely accept
the
> book. Indeed, I encourage
> > it. If the book is missing page numbers, or has running headers, or the
> table of contents is
> > jumbled, or the title page is missing (but the above info is still
> available), those are not bases
> > for rejecting the book.
> > >
> > > I predict that many volunteers will post follow-up messages regarding
> other things that they do,
> > and think that you should do, before approving a book. That is because
we
> have the most dedicated
> > volunteers in the world, and they go above and beyond the call of duty.
> But even they know that
> > while they may do additional work to improve a book that meets the above
> six requirements, they
> > should not reject a book that meets them all. So whatever the follow-up
> messages say, remember that
> > this message is the bottom line!
> > >
> > > I hope that clears it up for everyone!
> > >
> > > ________________________
> > >
> > > Jesse Fahnestock
> > > Collection Development Coordinator, Bookshare.org
> > > www.bookshare.org
> > >
> > > A Project of The Benetech Initiative - Technology Serving Humanity
> > > 480 S. California Ave., Suite 201
> > > Palo Alto, CA 94306-1609 USA
> > > (650)475-5440 x133
> > > (650) 475-1066 FAX
> > > jesse@xxxxxxxxxxxx
> > > www.benetech.org
> > >
> >
> >
>
>
>
Other related posts:
