* Automated internet Marriages lookup

[RogerF]

I'd like to pick the brains of those younger and wiser than me.

I have 11,000 families in my database. For the majority of the couples, there is a citation for their civil marriage. For a much smaller proportion there is, as well or instead, a citation for their church marriage. I've never used the FH internet lookup features.

What I'd ideally like is a mechanism of some sort which automatically:
1. Cycles through each family.
2. Where no church marriage is recorded, searches for one online.
3. If found, adds a appropriate citation in FH.

Any chance?

[tatewise]

Hi Roger.
Just to get a few points straight:-
By "11,000 families" do you mean 11,000 Family records as indicated by the Families: count bottom right of the FH Focus Window or Records Window, and therefore roughly double that number of Individual records?
By a "citation for their civil/church marriage" do you mean a Marriage fact Citation of a Source Record containing their Marriage Certificate/Parish Record transcript in Text From Source and/or a Media tab image?

The first thing to make clear is that Internet searches usually do NOT reveal GRO Marriage Certificates, they only provide GRO Index pages, which for popular names often offers several alternatives, especially as spouse names are not included in the earliest GRO Index pages. Parish Marriage Records are available in some Internet searches, but again popular names often offer several alternatives. If the date and/or place of marriage is uncertain, then the search results will probably offer many more alternatives.

Automatic mechanisms:

1. To cycle through families will require a Plugin, similar to Lookup Missing Census Facts.
2. To search online will require a Plugin, as above.
3. Adding a Citation is NOT feasible for the reasons given above; Marriage Certificates are not available via online searches, and the alternatives offered need manual review anyway. Even if only one record matches, the task of extracting the transcript and downloading an image is very complex and different for each website (Ancestry, FindMyPast, MyHeritage, etc.), and regularly changes as the website design evolves.

In the past I have considered a Lookup Missing Marriage Records Plugin, but if there are few Date or Place details, there will be a lot of alternatives to sift through manually. (Whereas Lookup Missing Census Facts knows each Census Date, and relies on the Date and Place of Birth plus names of close relatives to narrow the search.)

A Query would be quite straightforward to list all Family records that fail whatever Marriage Citation criteria you choose.

[Gowermick]

Roger,
Whilst it seems like a good idea, I think it would need a pretty serious plugin to achieve this. It would be fairly complicated however, as church records tend to be spread over many sites, with no one site covering all churches!

I think the best you can do, is produce a query which lists all marriages that have a civil registration, but are missing a church entry and search for the marriage yourself.

Mike

[RogerF]

Thank you, Mike and Mike

Mike Tate:

Yes, 11,000 Family records.

By a "citation for their civil/church marriage" I mean some form of FH reference to the internet-discovered record of the marriage; at this stage I'm more interested in whether the concept is viable, rather than the specifics of exactly how it's recorded.

Yes, a query could report candidate marriages, but 11,000 of them? no thanks!

Mike Gower:

I know it's a big ask, and to be honest I wasn't expecting it to be possible. You have both confirmed my expectation that what I'm suggesting is beyond the possibilties of what FH can accomplish.

Thank you both for supporting my suspicion that this can't be done. But just think how annoyed I've have been if I'd plodded through the manual lookup route, and THEN discovered that FH could have done it!

Thanks again, Roger

[tatewise]

Roger, let's explore the ideas a bit further, because with 11,000 records any amount of automation is likely to help.

Whether by Query or Plugin it is not necessary to review all 11,000 Families in one go.
They can be grouped into batches by various criteria: Related family groups, Relation Pool numbers, Record Id ranges, etc.
See the Lookup Missing Census Facts Plugin for the selection options it offers.

Your original posting said the majority of Family records already have a Marriage Citation.
So I assumed you only wanted to research the minority missing a Citation, and that is what your point 2. implies.
Therefore, the records listed by either a Query or Plugin would be much fewer than 11,000?

I do not understand your distinction between civil and church marriages - please explain.
Because of that, I am not clear about what online lookups you expected to achieve.

I have previously considered a Lookup Missing BMD Records Plugin:

• The Lookup Missing Census Facts chooses Census records by Country.
  The BMD variant could choose Birth/Baptism/Christening(BBC) or Marriage(Marr) or Death/Burial/Cremation(DBC) records by Country.
• The main task is to identify event Date and Place to narrow the online search:
  • That is quite easy for BBC & DBC because FH has EstimatedBirthDate() & EstimatedDeathDate() functions using fact Dates from the Individual, and close Relations if necessary, with an earliest & latest Date option.
    The Place details could be extracted from the individual, their parents, or siblings.
  • The Date range for Marr must be estimated by Plugin using spouse and 1st child EstimatedBirthDate().
    The Place details could be extracted from the spouses and children.

The most effective way of working would be to use Lookup Missing Census Facts first, and create the Census Citations, because that builds Family relationships with at least approximate Birth Dates & Places for everyone, but only covers around 1800 to 1940.
Then lookup BBC citations, then DBC citations, and finally Marr citations.

[RogerF]

Thanks for the positive reply, Mike.

Your original posting said the majority of Family records already have a Marriage Citation.
So I assumed you only wanted to research the minority missing a Citation, and that is what your point 2. implies.
Therefore, the records listed by either a Query or Plugin would be much fewer than 11,000?

I do not understand your distinction between civil and church marriages - please explain.
Because of that, I am not clear about what online lookups you expected to achieve.

The majority -- post 1837 -- have what I, probably incorrectly, term a civil marriage citation: that is, what you get from for example FreeBMD: Q3 1893, North Bierley, 9b, 411. A minority also or instead have what I term a church marriage, found in, for example "West Yorkshire, England, Church of England Baptisms, Marriages and Burials, 1512-1812", which usually gives the actual marriage date, and often the bride and groom's fathers.

What I would ideally like to do is look up every marriage where I have only the civil citation, to see if I can also find a church record.

[tatewise]

What you call a civil marriage citation is what I earlier called a GRO Index.

I presume you know that formal Marriage Registration was introduced in UK in 1837.
Each Parish Marriage Record became a GRO Marriage Certificate and its reference compiled into a quarterly GRO Index.
So that FreeBMD: Q3 1893, North Bierley, 9b, 411 is a reference to the 3rd Quarter 1893 GRO Index for Registration District North Bierley volume 9b page 411.
That same reference is what a Marriage lookup would find on Ancestry or FindMyPast, but their GRO Index entries are complete, whereas on FreeBMD they are incomplete.

With that GRO Index reference you can order a Marriage Certificate for £9.25 from GRO online.

In some cases it may be possible to find the original Parish Marriage Record online that contains the same details.
However, that only applies to Church Marriages, and not Civil Ceremonies held at a registrar office.
I am not sure how many such Parish Marriage Records actually exist online.
Checking my family history only found a handful of online post-1837 Parish Marriage Records.
Pre-1837 marriages did often find transcripts with various details, and sometimes images.

With a few tests on FMP, apart from confirming the Year exactly, having the GRO Index does not help much.
The search results still must be eye-balled to check if valid, and select a relevant one with any new details.
This is similar to Lookup Missing Census Facts Plugin that only composes online searches, but not any auto-citations.
However, even that amount of automation may well be beneficial.

The situation may be very different for other countries than the UK.

[brianlummis]

Although this may not apply to Roger's records, I record the Place of Marriage as "......... Registration District, County". If I subsequently find details of the exact place then this replaces the Registration District, and leave the original GRO Source and citation, as well as recording the new source.
Most online records giving the exact location tend to be by individual County Record Offices or a Family History Society in a County. So in Roger's example a query of marriages taking place in West Yorkshire can then be sorted to reveal all those with a place name that is a Registration District which narrows down the list to research.
Much of my research is based in Suffolk and I am waiting with baited breath for online images to be released and hours of work with Ancestral Sources to record them all!

[Valkrider]

Mike & Roger

Something that you may not be aware of is that churches can be calculated from the GRO reference. Work is ongoing by The Surname Society to put this data online on their website. http://surname-society.org and then select 'Church Marriage Finder'. There is also a similar project at http://www.marriage-locator.co.uk/ they have different data and new records are being added regularly to both, however, neither are yet anywhere near complete. They may give you some clues to the church records you need to search for though.

[tatewise]

From my experiments on FMP, the Place cannot be used to automatically filter the search, other than perhaps the County and Country.

Also the criteria for deciding an event is missing is more complex than for Census Events.
For a Census year the person must be alive on the Census Date and the Census Event must not exist for that Date - easy!
For say a Marriage the Marriage Event may not exist, or it exists without a simple unqualified Date, or it does have a specific Date but no Citation, or something like that. It gets more complex for the Birth/Baptism/Christening combo and Death/Burial/Cremation combo.

[AdrianBruce]

... What I would ideally like to do is look up every marriage where I have only the civil citation, to see if I can also find a church record.

What perturbs me is - where are you going to look for the parish records? We know there is a hinting process for FMP - that gives us a number of counties (e.g. Cheshire, Staffordshire, Shropshire and some Devon to take the ones I know). Ancestry would add in (say) some Somerset, Lancashire, Manchester and Liverpool (the last three being separate collections). There is plenty of other stuff on-line for Somerset across several other sites (mostly fixed HTML). Co. Durham has Durham Records Online. Scotland - well, you might use the index from FamilySearch, otherwise it's ScotlandsPeople.

So far as I know, apart from FMP, all those would have to be done using "screen scraping" (unless there is a hinting system for Ancestry?) and each site would need different scraping code to disentangle the PR information out of the HTML. I'd really love a push-the-button mechanism to interrogate the Somerset PRs but as there are so many sites, it'd probably be faster to do it manually.

Am I missing something about what's being pondered over?

[tatewise]

What is being pondered is an automated search similar to the Lookup Missing Census Facts Plugin.
(If you have not used that, then do so to understand more clearly.)
For users with missing BMD Citations it would lookup whatever is on offer primarily on Ancestry and FMP.
UK post-1837 would mostly find just GRO Index pages, but may include some Parish records.
UK pre-1837 includes a variety of records.
Other countries also have their own style of records.

Scrutinising the Records listed by the search would be manual just like those found by 'hints' and the Census Plugin. Similarly, the creation of Citations and Source records would be manual using drag & drop, or cut & paste, etc.

[StevieSteve]

If we park the 'automate the entry into FH' then I think the requirement can be boiled down to

FOR each marriage
Do I only have the GRO reference?
IF Yes
Look for a parish record

The way I enter my data, you could look for those marriages that only have a Quarter date format or that has a type of "MarriageRegister"

Then you would have lookup parameters of

Individual name
Spouse name
Date
Place (Registration District)

which could then be used to search the online sites though I wouldn't trust using the place

Now, if the marriage is in Middlesex that gives about an 80 year range where useful results are possible at the moment. Each county is obviously different.

Whether RogerF would find the continued manual element as sufficient benefit to him is open to question.

[RogerF]

This is looking more promising. Following up StevieSteve's thoughts, if those parameters could be used to search for Parish Marriages with the year exact and the couple's names exact or close, and produce a list of candidate marriages, it would be be very helpful. In fact, the year could be a supplied parameter to make the output more manageable.

[tatewise]

For a Plugin to be of general appeal it must cope with more than just Marriages with a GRO Index Citation, otherwise the effort creating the Plugin is not worthwhile ~ not for me anyway!

I outlined the criteria for initiating a lookup earlier, based on event Date and Citation details.

It should cope with no Citation at all, so that GRO Index entries can be found, and also UK pre-1837 Parish Rrecords, and other country records can be found.

Unless Stevie knows otherwise, the FMP (and probably Ancestry) search parameters do NOT allow only parish records to be chosen. That must be a human scrutinisation process.

If the event Date is a Quarter Date then the Year is an automatic search parameter with range +/-0.

Stevie's first criteria is "for each Marriage" which is OK if a Marriage Event already exists, but to be widely appealing a Plugin must cope with each Family record having no events, so the GRO Index can be found in the first place, or find any Parish Record.

The concept behind this type of Plugin is very similar to the FH 'hints', but instead of just the Focus Window family, any Individuals can be searched, and instead of all records, just selected classes of records are searched.

[AdrianBruce]

Bear in mind that if you're starting from a GRO marriage index entry, you don't know what the parish will be. Nantwich Registration District, for instance, covered a multiplicity of churches. You should, however, already have the county and if you have the exact year from the GRO entry, then there won't be many such marriages in that one year - even if John Smith is marrying Mary Jones in London....

[brianlummis]

Nantwich Registration District, for instance, covered a multiplicity of churches. You should, however, already have the county and if you have the exact year from the GRO entry,

Which is what I was implying in my earlier post.

It is unlikely that all the marriage transcriptions and/or images of parish registers will be available online for quite some time. Therefore the first thing to ascertain is which ones are available. Once you have that information you can narrow down your search to that county or area. I am not sure what other people record as Place when entering GRO Index information but I would guess that at the least it should include the county. If so a simple query will suffice.

Brian

[RogerF]

I'm well aware that there are a multiplicity of sites providing a local view of BMD in their domain; these sites are clearly outside the current remit of this thread. All that I'm hoping for is a scan of the major players -- Ancestry and FindMyPast for example -- in the hope that they may highlight some of the families I'm working with.

[AdrianBruce]

... All that I'm hoping for is a scan of the major players -- Ancestry and FindMyPast for example ...

OK Roger - eminently sensible. I'm sufficiently pedantic from my days of constructing IT requirements that I wouldn't restrict to the major players unless explicitly authorised!

[tatewise]

I have moved this to the Plugin Discussions Forum because that is the direction it is heading.

To have something more concrete to ponder I've built a prototype based on my Lookup Missing Census Facts Plugin.

See the attached Lookup Missing BMD Records Plugin Version 0.1 Date 16 Sep 2016.

It is far from the finished article, and currently only deals with the 1st of multiple Marriages.
It reviews each BMD Date and if existing, but not a simple full Day-Month-Year, then it assumes data is missing.
So that includes where only a Year exists, or a Quarter Date, or Date Range.
It uses FH Estimated Birth Dates to review Birth/Baptism/Christening Records.
It uses FH Estimated Death Dates to review Death/Burial/Cremation Records.
It uses Marriage Date or 1st Child Birth Date and Marriage Status to review Marriage Records.
It does not yet use Citations in its criteria for deciding if data is missing.

[Attachment deleted as now superseded by later versions.]

[tatewise]

See the attached Lookup Missing BMD Records Plugin Version 0.2 Date 17 Sep 2016.

This now deals with multiple Marriages correctly.
It now correctly filters on the Date as described for Version 0.1 but wasn't working correctly.
Now includes Place details but may need refinement to only include County &/or Country.

[Attachment deleted as now superseded by later versions.]

[tatewise]

See the attached Lookup Missing BMD Records Plugin Version 0.3 Date 19 Sep 2016.

It just keeps getting better, but I would appreciate some feedback if you have tried it.

Only uses last two parts of Place field, and prunes unwanted text.
When full Event Date, if no Citation, then treats as records missing.
Plus other minor refinements.

[Attachment deleted as now superseded by later versions.]

[jimlad68]

Bit late for reply, catching up after being away.

I've not done any automated lookups yet (wife does most of the research while I fail to keep up with the keying!) so I can't give any in depth response, however at first glance looks very useful and would certainly save "a lot of clicks"

Only thing I can see at present is various 'options/ exclusions' for things like, but not limited to, PLACe/ADDRess as not everyone uses the same structure, and in my case, not only do I reverse my PLACe fields (for sorting purposes), but I also use current addresses to tie in with GPS, google maps etc.

Another nice one Mike.

[tatewise]

See the attached Lookup Missing BMD Records Plugin Version 0.4 Date 05 Oct 2016.

This version adds:

• Help & Advice pages
• Option to exclude persons abroad at BMD Event dates
• Option to use First 2 parts or Last 2 parts of Event Place fields in Search criteria

[Attachment deleted as now superseded by later versions.]

[jbtapscott]

Mike - some feedback from my initial use of v0.4.

Firstly, some clarification of my data as this may well have some relevance to the output from the Plugin:
Places - I use a 5 column approach for P{laces data with, for GRO Index related records, the Registration District recorded in Columns 3 and 4 (Town/City and County) and then I have the Country in Column 5;
Dates - For Births / Marriages / Deaths using the GRO Index, I record the Date (from the GRO Index record) in the relevant Fact as the Quarter and Year (eg Q1 1896).

The Plugin options I had set were: Births / Baptisms, Ancestry - Disbabled, FMP - UK, All the Relations of a Chosen Person, Home / Abroad Unticked, Places - Last 2 Parts.

What I found was that most of the records in the Result Set (and thus the links to FMP) were ones for which I already had a Birth Fact in FH based on the GRO Index record from FMP so, basically, reporting information I already had. Have I misunderstood what the Plugin is trying to achieve / is there a need for some more Options (eg use of Quarter Dates rather than a "full" date)?