improving the interface for correcting record indices, additions, deletions, locales

LegacyUser

Vincent Broman said: I have submitted a lot of corrections to the record indexing already done, using pages like https://www.familysearch.org/ark:/619... . Several changes in the user interface could improve the work and the results.

1. It's not uncommon for given names to be in error because no given name is present in the record, or inversely a readable given name was not indexed. The same for surnames. The interface does not allow removing or creating given names, or surnames. Working around it, I create names like "NONE".

2. Similarly we need to add missing dates or remove illusory ones.

3. The locale chosen for dates should be either English or the language of the document, not the official language of the country eventually containing the place described. My ancestors were Swedes with Swedish names and documents, living in Vasa County. Just because their distant descendants pay taxes to Helsinki does not mean that they are Finnish. Finnish month names are a mystery.

4. Sometimes records are created for comments or ink blots, and sometimes real people are missed. Creating or Deleting records would help.

LegacyUser

Lundgren said: Thank you for your feedback.

I don't have solutions for your requests above, as I do not work in that area, the teams over that area can monitor this forum though.

However, adding names like "NONE" will create a searchable entry for the name NONE. I don't think that is really what you are after. "None" could be a name in another language, and you will be creating bad search results for someone else.

LegacyUser

Vincent Broman said: About the use of "NONE": it is obvious to me also that if there is no surname in the image, then the surname field should be blank. Same for given names or for dates.
Unfortunately, the user interface does not allow that to happen. A non-blank name can only be corrected to a non-blank name, and so forth. An illusory date can only be corrected to be another date. My suggestion to familysearch is to fix the software to handle this problem so that squiggles or comments that are not names but which got indexed as names can be removed. Similarly for the inverse problem of writings which got ignored/neglected but need to be added.

I hate FT records with personal names like "Unkn" or "?" as much as anyone, but I think in an index "NONE" is more useful and less confusing than an appropriate looking name which happens to be totally wrong.

LegacyUser

Vincent Broman said: As I read the two responses so far, I see I may not have made myself perfectly clear, that I am talking about correcting existing index entries, not about doing indexing or about creating FT records of persons.

LegacyUser

Lundgren said: Thank you for clarifying.

Correcting the names to "NONE" creates a name entry of None in the search results.

If you do a search for the surname of None, you will find that there are people in the the system that have real records that indicate their surname is None.

Here is a search:
https://www.familysearch.org/search/r...

Here is a sample record that indicates the Surname is None: https://www.familysearch.org/ark:/619...

Using Non for empty fields is equivalent to setting all names that should be empty to Broman. Anyone with that surname would not be happy to find the system more full of results that are clearly not correct.

You will probably also find records that have been incorrectly set to None as you have indicated you have done.

You are creating bad data by entering none. Until they come up with a way to remove empty names from the index, you should not change the data at all.

LegacyUser

Juli said: Yup, what he said.
However, the corrections feature will not allow leaving the field blank. Most types of punctuation are also disallowed, but a dash (-) works.

LegacyUser

Juli said: Tom, you appear not to have comprehended the basic topic.

LegacyUser

Juli said: Use a dash.

I just did a search and found that someone has gone through and changed a bunch of illegitimate children's fathers to "Ismeretlen Unknown" in the Hungarian baptismal indexes. This is Bad: it means that those entries will be cluttering up all sorts of wildcard searches.

LegacyUser

Tom Huber said: Actually, I have. I originally intended to report that the correction mechanism is far from complete. I believe the ultimate goal is to have all fields of a index from volunteers on FamilySearch to be correctable.

In the meantime, after looking at your response to the first reply, I now realize that the issue is not clearly set out in the original post.

From what I can gather, the idea is to be able to make an index field blank (delete the entry).

LegacyUser

Adrian Bruce said: For info: Re "no given name is present in the record" - a special case of this is where someone's complete name has been omitted from the original record, but an entry is created for them - so neither given nor family name is in the original.

Typically this applies to British baptism registers earlier than sort-of, vaguely, in round terms, 1800. In that case a child's name is present but probably only a father's name. The useful thing to to is create an index with one child and only one parent - normally that happens.

However, on occasions, someone in the indexing programme decides that because there was a mother, there should be a mother in the index with a name of "\" (I think). Since this serves no useful purpose, I think that it would be useful to delete entirely the index record for the "\" mother - not just the name, the entire index record for this persona. This would save the need to mess about merging the "\" mother with an existing mother with a real name.