Specify search form requirements

[PatReynolds]

Specification has two parts
a) a list of the fields which will be searchable
b) options for how the search interface will work, e.g.
i) a simple 'all the things you might want to search by are here on this page' interface
ii) a slightly more sophisticated version where there is a very simple name/place/date search, with an advanced search with the rest (perhaps accessed through a 'find' which takes you there if more than a certain number of records are found).
iii) an AI interface which might start with name (ticky box to restrict to testators), county and date, and if limit is reached, offer further options

[PatReynolds]

https://probatesearch.service.gov.uk/#wills starts with surname and date. The results for pre 1996 are just an image from which you have to pull relevant details to submit your request!

http://www.attrill.info/Attrill%20Family%20research%20Material/index.htmlhttp://www.attrill.info/Attrill%20Family%20research%20Material/Attrill%20wills%20&%20%20Probate/Attrill%20Probate%20Records.html has an interesting (anarchic) approach.

http://www.gingell.com/familytree/?page=probate just has a list either whole or filtered by single year, which can be ordered by first name, location or effects.

[PatReynolds]

http://www.nationalarchives.gov.uk/help-with-your-research/research-guides/wills-1384-1858/

[DeniseColbert]

Recent discussion in busy genealogy Facebook group about the lack of a national Probate facility that allows search by address.

[PatReynolds]

Search terms https://docs.google.com/document/d/1___qS9PCENrW2oRt4uC22bAfmfxu9MDBENTsHP_MhqI/edit?usp=sharing

[Captainkirkdawson]

It would help to have some examples of the documents being proposed to be indexed so we can see what may be there before going too far in defining the search itself

[PatReynolds]

@Captainkirkdawson the examples transcribed to 'gold standard' for machine learning to take place are in the FreePROBATE folder that I have just given you access too, and the bulk for machine learning to use are in a dropbox account (let me know if you need access).

But other wills are already being proposed for inclusion (via separate project, as it will need human transcription, rather than machine learning / correction). And we should therefore consider what fields might be needed from those other projects.
My initial thoughts - https://docs.google.com/document/d/1JwVls8udkVw8NFe7c1Y2WQRdsSIjxhVWa1exzP2kv90/edit?usp=sharing

[richpomfret]

Ask researchers (in newsletter & social media) what and how they would like to search wills.

[benwbrum]

@Captainkirkdawson Here is an example of the sort of document we're using as sources. I think that we can use these as analogues to burial records in FreeREG and marriage records as well (when it comes to witnesses): https://github.com/FreeUKGen/SummerOfCodeImages/blob/master/wills/images/1873_Probate_G_to_I/00028.jpg

[Captainkirkdawson]

Norfolk ones are

not as pretty.

[Captainkirkdawson]

I do not see including them as part of REG they are different animals; likely well suited to mongodb and its text search

[richpomfret]

@DeniseColbert to ask about required search fields in newsletter (include as a form).

[DeniseColbert]

Form sent out in newsletter. Email received via RQG re. https://transcribedwills.co.uk/Will-Finder - suggesting collaborating/partnering.

[richpomfret]

To review and then let this lead into creating the product specifications

[richpomfret]

@DeniseColbert did we get much feedback re: the form we sent out in the newsletter? If so, we will need to pencil in a meeting to review it.

[DeniseColbert]

Ben has already reviewed, will send link to responses in an email...

[richpomfret]

Converted to Epic and set up a story for reviewing search form feedback. We can now put this back into an Epic column.