News on Database errors project (12 June 2016)

Question

Analysis was done on data from June 12th 2016.

So in 7 days 20525 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

	5.6.	Projected	12.6.	Reduction	Delta%
101 Birth in future	236	237	237	0	0,10%
102 Death in future	272	273	239	34	12,59%
103 Death before birth	14087	14161	13867	294	2,08%
104 Too old	6557	6591	6400	191	2,91%
105 Duplicate sibling	2365	2377	1955	422	17,77%
106 Duplicates between bigtree and unconnected	2664	2678	2502	176	6,57%
107 Full name in UPPERCASE	2997	3013	2878	135	4,47%
108 Full name in lowercase	3118	3134	3114	20	0,65%
109 Profile should be open (birth date)	18604	18702	18379	323	1,73%
110 Profile should be open (death date)	1571	1579	1521	58	3,69%
201 Father is self	109	110	104	6	5,09%
202 Parents are same	71	71	79	-8	-10,69%
203 Father is Female	6243	6276	6122	154	2,45%
204 Father has no Gender	898	903	685	218	24,12%
205 Father is too young or not born	42799	43024	42544	480	1,12%
206 Father is too old	6577	6612	6510	102	1,54%
207 Father is also a child	364	366	360	6	1,62%
208 Father is also a spouse	65	65	62	3	5,11%
209 Father is also a sibling	2828	2843	2701	142	4,99%
210 Father was dead before birth	42179	42401	41757	644	1,52%
211 Duplicate sibling by father	4542	4566	4515	51	1,11%
301 Mother is self	3	3	4	-1	-32,64%
303 Mother is Male	6626	6661	6469	192	2,88%
304 Mother has no Gender	867	872	867	5	0,52%
305 Mother too young or not born	54490	54775	54015	760	1,39%
306 Mother is too old	5564	5593	5511	82	1,47%
307 Mother is also a child	8	8	8	0	0,52%
308 Mother is also a spouse	989	994	914	80	8,06%
309 Mother is also a sibling	231	232	237	-5	-2,06%
310 Mother was dead before birth	48198	48450	47557	893	1,84%
311 Duplicate sibling by mother	1644	1653	1601	52	3,12%
401 Spouse is self	2	2	2	0	0,50%
402 Unknown gender of spouse	1681	1689	1562	127	7,54%
403 Single sex marriage	1406	1413	1105	308	21,80%
404 Marriage before birth	10925	10980	10717	263	2,40%
405 Married too old	2750	2764	2694	70	2,53%
406 Marriage after death	13268	13335	13046	289	2,17%
407 Death too old after Marriage	1420	1427	1303	124	8,70%
408 Multiple marriages on same day	9853	9903	9644	259	2,61%
409 Marriage to duplicate person	31076	31232	30637	595	1,91%
501 Wrong male gender	8291	8335	7617	718	8,61%
502 Missing male gender	73695	74083	71185	2898	3,91%
503 Probably wrong male gender	6095	6127	6017	110	1,80%
504 Probably missing male gender	35528	35715	35209	506	1,42%
505 Wrong female gender	8987	9034	8715	319	3,53%
506 Missing female gender	59846	60161	58779	1382	2,30%
507 Probably wrong female gender	4985	5011	4910	101	2,02%
508 Probably missing female gender	28538	28688	28497	191	0,67%
509 Missing gender	96461	96968	96542	426	0,44%
510 Unique name without gender	23171	23293	22851	442	1,90%
511 Unique name (spelling)	237250	238498	237321	1177	0,49%
512 Separators in first name	68136	68494	67608	886	1,29%
601 Unknown birth location	9026	9080	8970	110	1,22%
604 Birth location too short	13787	13870	13754	116	0,84%
605 Number in birth location	1742	1752	1621	131	7,50%
606 Bogus birth location	11	11	4	7	63,85%
631 Unknown death location	16148	16245	15927	318	1,96%
632 Y death location	5529	5562	4936	626	11,26%
634 Death location too short	15752	15847	14575	1272	8,03%
635 Number in death location	1045	1051	458	593	56,43%
636 Bogus death location	939	945	643	302	31,93%
661 Unknown marriage location	1349	1357	1272	85	6,27%
664 Marriage location too short	2789	2806	2825	-19	-0,68%
665 Number in marriage location	14	14	14	0	0,60%
901 Unconnected empty public profiles	35194	35379	35192	187	0,53%
902 Unconnected empty open profiles	17489	17581	17486	95	0,54%
Total	1121944	1127876	1107352	20525	1,82%

 

Comments

This is so awesome ! Keep up the great work and thank you all for correcting !

---

Thanks to you and everyone who is working these errors....

Keep up the great work!

---

Yes, this work and this table strikes me with AWE. It's in-credible what tech has wrought and in the hands of smart caring people.

I have found my errors list, and in some cases examples abound that should not be corrected:

For example: a) the unusual name IS the person's name. (b). no other name exists in records of Jewish persons or their precise days of B & D.

MY QUESTION NOW: Is there a way to copy the errors table into a table that's a Word document (.docx) so it's easier to scroll and check and change? OR How are you performing the mechanics of this process in terms of going between sets of data on different sites and making decisions? Is there a clear way through such a clotted river? Or is the slogging just, well, slogging? (yes, the last is probably the "only way to go.")

 

 

---

You can just save the page (Save as, or Save Page as, or whatever your browser calls it) and then open that file in MS Word or any similar program.

I never need to do that, though; the browser lets me know which ones I've clicked on (purple instead of blue) and that's about all I need.  If you just do a few at a time, you can also click Temp Hide and those will be gone the next time you reload the error list page.

---

Roberta - I have the browser add-on called STYLIST which lets me change the color of websites I've visited to red so that I don't have to try to remember which errors I've fixed.

---

That sound really helpful, thanks Deborah !

---

Roberta,

 there is no hurry in repairing these errors. Start with the easiest ones, and continue with the ones that need more research. When you correct one error, you can click Hide for 30 days and error will be removed from the list (Refresh the page). Errors are recalculated each monday and corrected errors will no longer be present. 

---

I'm on a Mac, so does your idea of using "the browser add-on called STYLIST" work on Macs? I want to avoid frustrations and confusions.

---

Alesh, Thank you. Your comment is consoling. I'll try that. The first error noted on my list is not an error but a variant spelling of a first name. That one is correct as it is.

---

Rebecca, they have a version of Stylist add-on for almost all browsers - a twinge of jealousy for your Mac :) - just google the add-on Stylist

---

Stylish (with H) is the popular one at least for Firefox.but IMO she doesn't need it.

---

IMO??

---

In My Opinion

---

In My Opinion - Mikey - maybe you could give her clear directions on how to change the color of her visited websites on MAC - not my forte :)

Answer 1

I would like to congratulate you on the work it took to get that database up and running, well done

Answer 2

It appears my corrected errors were not updated this week?

See Seaton-582 Death date corrected from 1864 to 1964, but error tool still showing 1864.

Comments

Wait a few minutes. Lists are updated. Online errors are preparing.

---

Thank you, you do sleep, yes?

---

It is up. It is 4.00 pm here, so I don't sleep now.

---

Michael, it appears you've corrected the death of the child [Seaton-582], but not the death date of the father [Seaton-581].

---

Yes that is true. It is the error of the daughter that hadn't updated. But now has.  What the father error pointed out is that I need a source for Hester being the daughter of William which I am currently investigating.

Answer 3

This profile shows as open in the list of 902 Unconnected empty profiles. But when you go to it the profile is closed.

Abercrombie-258

Comments

Dawn,

The profile was edited this morning by Jillaine, and update of error report is next monday, so that probably explains the difference. There currently is no profile manager so you can adopt it if you like.

Answer 4

Should the system check for married couples with the same LNAB?  Sometimes it happens, but most of them will be Mrs John Smiths.

Of course the next question then would be, whether the maiden name is really Unknown, or whether somebody never got round to looking.

Comments

RJ, I am at this time working on a line, Blackwelder, where there is a lot of marriages between people of the same LNAB. If I were inundated by a bunch or requests by others to check for "errors"that are not really errors it would make working on WikiTree  more pain than it is worth. I find it difficult to work on doing quality research as it is and adding to my workload because some others do not think I am doing what I should is just not worth it. Gender, locations and even dates should be in the project but to keep adding supposed errors strikes me as trying to keep the project going even when the number of profiles being corrected is making the total number on the error list decrease.

---

Actually, it happens enough to make it a total nuisance if you put in this as an error category.  You should be looking at whether there are valid sources for the data when this question comes up.

---

OK - husband and wife have same LNAB and wife has no father

---

RJ, In the family I mentioned they were distant cousins and she did have a father but even if no father was listed it would still not be an error because there is documentation that they were cousins. Another famous example is Franklin Roosevelt and his wife Eleanore Roosevelt. Lets focus on the errors that need to be fixed and not try to find things to make us look worse that may not be errors.

---

With more than 3 million errors on the current report, I think we have plenty to keep all of us busy for now.

---

Sorry, I now see this had already been suggested and is on the to-do list.

---

Agree.  Lets fix the egregious CLEARLY WRONG errors ( e.g.,4 year old women and 90 year old women giving birth) before we worry about things that MIGHT be errors.

---

There are 200,000 date-conflicted parents, and mostly they need research, and they'll take years and years and years.

Picking off the low-hanging fruit first is the efficient way to improve the site.

All the Mrs John Smiths just make WikiTree look like a clone of the worst LDS junk.

Going through the list marking all the genuine cases as false errors can be done very quickly and easily.  Then they disappear for good.

---

maybe the error shouldn't be defined as partners with the same surname but partners with both given and surname the same... that should flag all the John Smith/Mrs. John Smith pairs without tagging the John Smiths who married someone also born Smith.

Answer 5

Limit tree errors to blood relatives

In tree errors, I added Limit to blood relatives: CheckBox on http://www.sdms.si:92/wikitree/ShowErrors.htm

That means that relatives are limited to blood relatives. That means All Ancestors of start profile and all their descendants.

Comments

Wonderful!!!!!!

---

Thank you!

---

Wow!! That helps SO much!! Thanks!