News on Database errors project (5 June 2016)

+27 votes
697 views

Analysis was done on data from June 5th 2016.

So in 7 days 18209 errors were corrected by my estimation. Great job.

Have fun correcting errors.

You can also join the project here: http://www.wikitree.com/wiki/Project:Database_Errors

 

  29.5. Projected 5.6. Reduction  Delta%
101 Birth in future 242 243 236 7 2,90%
102 Death in future 307 308 272 36 11,79%
103 Death before birth 14153 14215 14087 128 0,90%
104 Too old 6621 6650 6557 93 1,40%
105 Duplicate sibling 2687 2699 2365 334 12,37%
106 Duplicates between bigtree and unconnected 2789 2801 2664 137 4,90%
107 Full name in UPPERCASE 3088 3101 2997 104 3,37%
108 Full name in lowercase 3193 3207 3118 89 2,77%
109 Profile should be open (birth date) 18738 18820 18604 216 1,15%
110 Profile should be open (death date) 1664 1671 1571 100 6,00%
201 Father is self 112 112 109 3 3,11%
202 Parents are same 79 79 71 8 10,52%
203 Father is Female 6376 6404 6243 161 2,52%
204 Father has no Gender 954 958 898 60 6,28%
205 Father is too young or not born 42740 42928 42799 129 0,30%
206 Father is too old 6648 6677 6577 100 1,50%
207 Father is also a child 375 377 364 13 3,36%
208 Father is also a spouse 208 209 65 144 68,89%
209 Father is also a sibling 2967 2980 2828 152 5,10%
210 Father was dead before birth 42118 42304 42179 125 0,29%
301 Mother is self 8 8 3 5 62,66%
303 Mother is Male 7333 7365 6626 739 10,03%
304 Mother has no Gender 1074 1079 867 212 19,62%
305 Mother too young or not born 54347 54582 54490 92 0,17%
306 Mother is too old 5606 5630 5564 66 1,18%
307 Mother is also a child 12 12 8 4 33,62%
308 Mother is also a spouse 1186 1191 989 202 16,97%
309 Mother is also a sibling 351 353 231 122 34,47%
310 Mother was dead before birth 48175 48383 48198 185 0,38%
401 Spouse is self 2 2 2 0 0,44%
402 Unknown gender of spouse 1742 1750 1681 69 3,92%
403 Single sex marriage 2156 2165 1406 759 35,07%
404 Marriage before birth 11152 11201 10925 276 2,46%
405 Married too old 2760 2772 2750 22 0,80%
406 Marriage after death 13271 13329 13268 61 0,46%
407 Death too old after Marriage 1583 1590 1420 170 10,69%
408 Multiple marriages on same day 10011 10055 9853 202 2,01%
409 Marriage to duplicate person 31338 31476 31076 400 1,27%
501 Wrong male gender 8515 8552 8291 261 3,05%
502 Missing male gender 74330 74655 73695 960 1,29%
503 Probably wrong male gender 6389 6417 6095 322 5,02%
504 Probably missing male gender 35693 35849 35528 321 0,90%
505 Wrong female gender 9699 9741 8987 754 7,74%
506 Missing female gender 59697 59958 59846 112 0,19%
507 Probably wrong female gender 5575 5599 4985 614 10,97%
508 Probably missing female gender 30243 30375 28538 1837 6,05%
509 Missing gender 95072 95487 96461 -974 -1,02%
510 Unique name without gender 23373 23475 23171 304 1,30%
511 Unique name (spelling) 293826 295110 293941 1169 0,40%
512 Separators in first name 68177 68475 68136 339 0,49%
601 Unknown birth location 9003 9050 9026 24 0,26%
604 Birth location too short 13796 13868 13787 81 0,58%
605 Number in birth location 1764 1773 1742 31 1,76%
606 Bogus birth location 58 58 11 47 81,13%
631 Unknown death location 16198 16282 16148 134 0,82%
632 Y death location 5662 5691 5529 162 2,85%
634 Death location too short 16033 16116 15752 364 2,26%
635 Number in death location 1239 1245 1045 200 16,09%
636 Bogus death location 980 985 939 46 4,68%
661 Unknown marriage location 1350 1357 1349 8 0,59%
664 Marriage location too short 2802 2817 2789 28 0,98%
665 Number in marriage location 11 11 14 -3 -26,62%
901 Unconnected empty public profiles 35432 35587 35194 393 1,10%
902 Unconected empty open profiles 17235 17310 17489 -179 -1,03%
Total 1180318 1185533,466 1171499 13084 1,10%
 

 

in The Tree House by Aleš Trtnik G2G6 Pilot (797k points)
edited by Maggie N.

6 Answers

+7 votes
 
Best answer

Added Error 211, 311

Added error 211 Duplicate sibling by father and 311 Duplicate sibling by mother, that lists profiles wits same FullName, birth and death date and one parent. It is similar to error 105 Duplicate sibling.

Errors Total 0001-1499 1500-1699 1700-1799 1800-1899 1900-1999
211 Duplicate sibling by father 4542 20 360 1086 2634 442
311 Duplicate sibling by mother 1644 16 128 378 952 170

Updated Error 511

When checking for uniqueness of a name, it is also checked against last names, so most of -son names and Latnames as middle name are now not reported as an error.

by Aleš Trtnik G2G6 Pilot (797k points)
selected by Living Anonymous
Can you explain errors 211 and 311? I don't understand the exact meaning.

Is it being named after their parent? Or what?

Jon

Here is more detailed explanation of the error 

http://www.wikitree.com/wiki/Space:DBE_211

http://www.wikitree.com/wiki/Space:DBE_311

Here are siblings with same name, birth and death dates and same father. Mother is different, but this two siblings are probably the same person. Also their mothers are probably duplicates.

+7 votes
Have you taken false errors into consideration?  Yes, they were corrections of a sort, but they wouldn't have been a error in the first place, and so I'm not sure what should be done with them.  Perhaps a way to just put the total of false errors would be helpful in how many real corrections were made.
by Dave Dardinger G2G6 Pilot (438k points)
There are 2500 false errors from beginning of the project. 2000 of them are for error 511. But i don't think all of them are correct. So it is very few of them except error 511.
+11 votes

Updated error lists

Prepared error lists are reduced in size to show 2000 instead of 5000 errors per page. Also sort order in lists is changed, On beginning are open profiles, that are easily edited and followed by more protected ones.

by Aleš Trtnik G2G6 Pilot (797k points)
Just noticed this. Nice change. Filtering on 60 was a bit awkward
When I was going through 502 list, I noticed that you open many profiles, that you cannot edit. So I decided to put opened profiles first. When all of them will be corrected, protected ones will remain, since they need different approach.
+6 votes
What do you think about inserting a separator line (---- in wiki, or <hr /> in html) after each error?  I have a hard time following across all the white space to the Temp Hide link. Not made any easier because it wraps to two lines.  Or, since the errors are in table format, you could just enable table border if that won't upset anyone.
by Living Anonymous G2G6 Mach 5 (51.2k points)
In going through the date errors, I see what appears to be deliberate attempts to set dates in the future, not just typos.  I think it would be worth an announcement or something official asking all members to please check their profiles for valid dates and explain they can choose "about/uncertain but non-living" rather than using false dates.

I also commented on the problem with long rows and find the matching temp hide

My suggestions

  1. Have CSS style so the row with mouse over is highlighted ==> Aleš checked and it was not a out of the box solution as they have CSS style on every cell that overrides...
     
  2. That we have a running number to the left and to the right ==> you can see that row 445 has a matching 445
     
  3. Change column for Temp hide and have it more to the left
Done.
Much better.  Thanks.
+7 votes

Updated Name analyses

With name occurrences check, it check for multi word names also each name http://www.sdms.si:92/wikitree/ShowFirstNames.htm. Automatic link to this analyse is added to each error.

by Aleš Trtnik G2G6 Pilot (797k points)
I am unable to access my Error Report at present to see any changes. Is their a problem at the moment?
It should work now. I am not sure what happened, but server got stuck. Restart helped. It was a huge increase in requests on my server today at noon (not sure why). I will monitor behavior more closely.
Thanks Ales;

I had also restarted my computer and all working OK now.
Today at noon news mailing list was sent and I am getting 1000s of requests.
+4 votes
Honestly, a significant number of these "errors" wouldn't exist if the database were created correctly - i.e. not allowing the creation of obvious errors.  But, bit by bit, the special-interest politically-correct woosificatoon crew has bullied its way into forcing the IT staff to add the option of allowing mothers and fathers to be "non-biological" when compared to their children, allowing fathers to be flagged as of female sex and mothers to be flagged as male.  

How about correcting the database and disallowing contributors to add altered reality, non-information that is clearly untrue. If a profile is identified as a mother, it should not have an option to identify as a male. Period. It should not have Y DNA tests attached to it - ever. If you have Y DNA results for a profile the subject is male and cannot be a mother regardless of how he dresses, feels or mutilates himself during life so there is NO need to have the option/possibility of creating such . The reciprocal is true for fathers. With two X chromosomes the subject is female, cannot be a father and shoukd have no option to ever be identified as male.

343,106 errors associated with misrepresented gender flags are reported on this report. If common sense limits corresponding to biology were in place on the database they wouldn't have to be corrected because they couldn't have been created.  Those limits aren't in this database, and never will be, because someone will complain that feelings are getting hurt, that profiles with red "Privacy" levels should be allowed to be configured any way the PM wants because they aren't here for public consumption (never mind that one day, 200+ years from now,  everyone on this site will surely be just as dead as our ancestors and a new generation will have to spend the time and energy to correct garbage profiles that become Open) - and no one will have the chutspa to just say "no".
by Michele Britton G2G6 Mach 2 (20.4k points)
edited by Michele Britton

Related questions

+34 votes
9 answers
+20 votes
1 answer
+33 votes
5 answers
+15 votes
0 answers
+20 votes
3 answers
+23 votes
3 answers
+20 votes
2 answers
+24 votes
4 answers
+23 votes
3 answers
+23 votes
5 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...