FREEBMD-DISCUSS-L Archives

Archiver > FREEBMD-DISCUSS > 2004-11 > 1101302145


From: Archer Barrie <>
Subject: RE: Project Update
Date: Wed, 24 Nov 2004 13:15:45 -0000


> At the moment the algorithm for calculating page ranges seems
> to require a
> minimum number of records.
>
> Thus districts with small populations don't get a mention.

At least one page number mentioned 5 times.

> Looking at the page range for 11b (North Wales) in one of the
> quarters where
> our syndicate has completed virtually every page, I can see
> that there are
> no ranges for Newcastle E., Lampeter, Aberayron, Aberystwith, Builth
> ,Crickhowell, Hay, Knighton, Machynlleth, Newtown, Forden, Holywell,
> Hawarden, Ruthin, St. Asaph, Llanrwst, Corwen, or Bangor.

If you could tell me which quarters and events I will look into it. We have
already discovered some tweaks that can be done to the algorithm.

> Having said that, I am a little surprised that there is no
> link from the
> Transcribers' Page -Volunteer Support and Resources - or from the
> Transcribers' Knowledge Base- to the page ranges that are on
> the Information
> pages if you know where to find them.

Getting links in the right place is a nightmare. Too many and it is
difficult to navigate, too few and people can't find the information. If you
would like to propose where the links should go we can look into it.

> I don't know how the new facility will work, but presumably
> the page number
> in a record found on a search is checked against the relevant
> page range.

Yes

> You also wrote 'We will also
> be using the information to identify files that have the
> wrong quarter/event (i.e. if most of the pages in a file are
> wrong it is probably the quarter/event that is wrong).'
>
> Presumably this process will check all files and possibly
> generate (at least
> transiently) a list of all entries outside page ranges. Might
> it be possible
> to make such lists available to Syndicate co-ordinators of
> the quarters
> concerned? It is quite laborious generating such lists oneself.

It doesn't produce a list, even transiently, it merely counts the
mismatches.

What form would such a list take?

Barrie


This thread: