* Individuals name filter suggestion

[ississi]

Hi, like lots of people I have lots of relatives with similar names (e.g. Thomas Lee ), who are sometimes differentiated by their middle name. As it happens, this is particularly true with my Scottish relatives - I have a lot of Gordon HAYs.
However, if I go into their initials I have GEH, GGH, GGWH, another GEH but with a different middle name, etc.

The name filter on the Individuals tab can find all GH's (last name H, first name G), including, very nicely, spouses despite their different maiden names, who could be subsequently listed as GH once married ( ), and those with GH as a subset of their initials, eg it finds Robert Gordon H, but it doesn't work for a search for GEH (last name H, first name "G E" - 'g' space 'e').

Would other people like it to do this?
And, if we were extremely fortunate, cope with eg: a search for last name H, first names G Edmond, vs searching for H, G Edward? (a combo of words and initials)?

Let me know what you think, or if you already have a way to do this , or if it would be way too horrible to try to implement in FH?

[Vyger]

Hi ississi,

This has always been a difficulty for the genealogy researcher, one of my Surnames derived from Scotland appears in almost 20 spelling variations. The challenge has been discussed before relatively recently so firstly please read the previous discussion on the link below and feel free to add your comments.

viewtopic.php?t=21168

A Wish List Item was created out of that previous discussion, please visit the item on the Wish List, vote on the proposal and feel free to add your own comments, you will need to be logged in to do so.

https://www.fhug.org.uk/wishlist/wldisp ... lwlref=633

Jackson

[Gary_G]

Jackson;

Frankly; trying to find a way to filter such a range of spellings is likely not possible, but there is likely a better way to solve the issue.

It would be nice to see FH7 do something similar to what is done for places and have the original and a "standardized" or "normalized" spelling. This approach has been used by Genealogie Quebec to facilitate searching the plethora of variations on French-Canadian surnames and the additional complication introduced by the "dit" variants. Implementing a similar capability would also accommodate showing and searching names in both the original alphabet (ie. Cyrillic) and the Roman alphabet.

[Vyger]

Gary,

Did you view the other discussion I linked?

I have seen good solutions to this problem, some of which I pushed for myself and that involves wildcards. In the other thread I show examples.

Since this thread is essentially a duplicate perhaps you could continue and add you experiences there.

Wildcards work well for me but I don't have much experience in my file of other country spelling standards and practices.

[Gary_G]

Jackson;

I did view the noted post, but it seemed focussed on an entirely different approach; that of parameterized searches. To add my suggestion to that thread would not be appropriate; especially as there has been a wish-list item generated based upon the existing content. That said; I think my suggestion could be the basis for a new wish-list item and that it would likely meet the need to be able to search for names that tend to have variants.

[ississi]

Hi ississi,

This has always been a difficulty for the genealogy researcher, one of my Surnames derived from Scotland appears in almost 20 spelling variations. The challenge has been discussed before relatively recently so firstly please read the previous discussion on the link below and feel free to add your comments.

viewtopic.php?t=21168

A Wish List Item was created out of that previous discussion, please visit the item on the Wish List, vote on the proposal and feel free to add your own comments, you will need to be logged in to do so.

https://www.fhug.org.uk/wishlist/wldisp ... lwlref=633

Jackson

Hi, yes, I did read it, but you seem to be discussing something totally different to me?

I'm talking about filtering on initials eg:

G E H would match George Edward Hay and George Edmund Hay

no wildcards needed, nothing about hyphenation.

[Vyger]

Hi, possibly my comments on the other thread we're not comprehensive enough for the purpose;

G%E%HAY% will find
Gordon Edward Hay
George Edmund Hays
Etc.

GO%EDW%HAY% will find more specifically
Gordon Edward Hay
Godfrey Edwind Hays
Etc.

Can you explain what additional search power you would require and how it is not covered by the proposed windcards?

[tatewise]

I'm not sure I fully understand how those wildcards work.
In [Wish List 633] Name Filtering to handle inconsistent entry or miss slashing of names (21168) you say that the wildcard "%" covers any number of characters. Also, your examples suggest the filters are case-insensitive.

So would G%E%HAY% find:
Grahame Shiphay Mills
Ginetta Hayley Jones
Agnetta Shayman

Perhaps you meant to use G% E% HAY% and GO% EDW% HAY% so the letters must start a word?

[Vyger]

Mike, as with any wildcarding there needs to a bit of thought and possible experimentation from the user to achieve broad or precise results.

All I can say is I use this to very good effect in a competitor program in a single field entry, not two or more fields as currently in FH. I have found it very useful on Scottish and Irish name variations where Mc or Mac might be in play and it has very good results with Scandic name patterns.

Regarding families, I can also enter Henry,Edith to filter only those partnerships. What I commonly refer to as SMART searching can be applied to Sources, Places, Research Logs etc and covers many fields rather than one so a very useful improvement in any program using AND logic.

As regards case sensitivity, I would need to check but I don't specifically remember so. That very point as an option could further help isolate propercase lead name characters.

Many users could quickly arrive at a satisfactory wildcard string, even those with programming experience may have to tailor things a little. As I am sure you are aware these are standard SQL wildcards so perhaps best to include an extract from the competitor help file to clarify what they offer in simplistic terms although in reality multiple combinations of wildcard characters can be used.


"..... will use the following wildcards in the Sidebar Index tab and People List search boxes.

•The underscore _ wildcard matches any single character.
Example: SM_TH, will return results for SMITH and SMYTH families

•The percent sign % wildcard matches any sequence of zero or more characters.
Example: Be%t will return results Bennet and Bentley"

[tatewise]

Thank you for your reply but it does not answer either of my specific questions.

If case sensitivity was an option and was enabled then:

1. G%E%HAY% will NOT find Gordon Edward Hay or George Edmund Hays because HAY does not match.
2. G%E%Hay% would work as long as the names did each start with upper-case letters.

[Vyger]

If case sensitivity was an option and was enabled then:

1. G%E%HAY% will NOT find Gordon Edward Hay or George Edmund Hays because HAY does not match.
2. G%E%Hay% would work as long as the names did each start with upper-case letters.

Firstly, let met me clarify the question above and confirm the competitor software offering is not case sensitive in this respect.

The name filter on the Individuals tab can find all GH's (last name H, first name G), including, very nicely, spouses despite their different maiden names, who could be subsequently listed as GH once married ( ), and those with GH as a subset of their initials, eg it finds Robert Gordon H, but it doesn't work for a search for GEH (last name H, first name "G E" - 'g' space 'e').

Allow me to point out that the example given above by ississi of 'g' space 'e' is incorrect formatting and you are correct FH will not currently return a match. Correct formatting is for any abbreviated text to be immediately followed by a period to denote the abbreviation and FH does provide a return for correctly formatted entries, see below.

wildcard-1.png (23.32 KiB) Viewed 260 times

Incidentally, my previous software had a feature called NameClean which would scan the project, report all suspected Name formatting issues, suggest corrections and allow for the selective action on those entries.

Would other people like it to do this?
And, if we were extremely fortunate, cope with eg: a search for last name H, first names G Edmond, vs searching for H, G Edward? (a combo of words and initials)?

For a concise example I have changed a Hayes in my database to match the forename of 'G Edmund' even though this is incorrect punctuation, the image below shows some of my Haye% entries.

wildcard-2.png (26.31 KiB) Viewed 260 times

Allow me to point out that if the name was correctly formatted as 'G. Edmund' (including the period) Family Historian will report it with simple 'g. ed' in the forename, without the period after 'G' it will not.

Applying the wildcard single comma delimited string below in the competitor software will return the entry regardless of the period being correctly entered or omitted.

hay%,g%edm

wildcard-3.png (13.28 KiB) Viewed 260 times

As a footnote Mike Tate raised the important point of case sensitivity. If CP were ever to implement such a single field comma delimited SMART search the optional inclusion of case sensitivity could add another useful criteria when using the % wildcard.

Name Filtering to handle inconsistent entry or miss slashing of names has been created and covers any of the needs previously outlined here. I would encourage logging in and voting on it if you want improved search/filtering functionality.