I have always found the Smart Matching system great, but limited.
I can go along and start adding new records based on confirmed matches and then at some point the system decides that a newly added match that came from some other tree no longer matches at which point I can not longer add preceding generations from that match.
Also, it would seem obvious to me, that if I add someone and there are more generations available (certainly preceeding) then why doesn't the system simply give me the option of importing all the succeeding generations from the tree that gave me the match! Clearly if I make a definate match, then everything else should match also, so give the choice to just import it all. It would make this much easier to build out.
Finally I encounter numerous trees where the builders neglect to mark people dead that are clearly dead - born in the 1800's and before. It shouldn't be too difficult for the software to determine by place in the tree and surrounding information that someone is clearly dead. This is getting in the way of the smart matching merging algorythms. Esp adding superfolous records.
I have just recently startet my own heritage-tree, and far out in the tree I have some SmartMatches, and the people have a lot more people that i didn't. Now I have tried to copy a lot of the persons from the other trees, but there is allmost never any notification of the origin of the name, death, birth an so on.
Which boils down to - if ONE person make an error in the tree, a lot of people would/could copy that error and in no time, there are 17 different trees with the same error. But every 17 people have a SmartMatch on the false name/person.
Does enough wrongs make a right? Do people not tend to tell where they have gotten the heritage-info from? Without a reference, it's allmost worthless because of the above mentioned; a lot of people think that they have the right name because they have 17 SmartMatches on the one person - but if it was an error in the beginning, then it's an error all the way through!
How to avoid that a lot of errors makes it seem right because of the SmartMatching?
This excellent thread highlights the main problem with SmartMatches: most people on MH are hobbyists, not genealogists, and as a result they make well-meaning but unfortunate choices. The terrible lack of accuracy is the reason I have chosen to totally ignore SM.
Here's my take on smart matches. Most of them are in fact "Dumb Matches" . Sugested matches are useless when the info being compared is in fact private, what was the point? My other gripe is that some/ a lot of my ancestors only had one name but this site has to create "Unknown Bill" or "Bill Unknown" for a person only called bill. So I have a lot of Unrelated people now related by the addition of "Unknown". I have been asked by Relatives to accept smart matches. When "Dumb Matches" and "Unknown" are handled better then I will be happy to become a finacial member. Untill then I am happy to use the site for what it is for my personal use and not for everyone elses benefit.
Sorry, not sure if this is the correct place to rage against the storm.
Yes, but the smart match will not show you the reference, even if their is a reference. The only way is to ask the site manager if you can be invited for a short time and review. I have a cousin who documents everything and I am a site member, and I can't even see his source, which I know he has. Turn off your smart matching and just do your own research, because some people will not share their hard work and research with strangers nor should they have too, after all they are paying for their a very healthly yearly premium, and have a right to block this from you.
I think I'll have to skip the smart-matches. Unfortunate i have allready added 150 people from smart-matches. I thought it was fun at the time. Some of them can stay, because they have sourcereference in the notes or elsewhere in the text.
The Idea to be able to flag errors on data is a great idea if it was public data as an article in a wiki. Here we have "private" family trees made public on MyHeritage. I'm not sure if I want to have public comments on my data about errors someone thinks exists due to his/her belief due to outdated sources. I think there would be a lot of hard discussions to try to make someone to back of their opinion (mine included). This I dont want to show in my tree. And how shall flags be handled if there has come to some sort of solution (correction or not)? Shall the flag(s) be removed, if so by whom? The flags could automatically be removed if there is a correction but if a correction is not needed a manual removal would be needed.
If the flag (with comments) would be only for the manager to see, then I think I could accept it since I could ignore it if I feel it is unwarrented. Problem is when the flags keep on coming from tree owners who has the data wrong.
I have difficulties today to have time to take care of all false SMs and to also take care of false error flagging would make me leave MH.
I agree that it is frustrating to see all wrong data and it realy itch in my fingers to point it out. Sometimes I get message about a confirmed match where the person in question has a difference in birth with several centuries. Managers that add children where the parents are born 200 years after the children. All probably due to a smart smart match they beleive is true without comparing the data and dont notice the differences in existing birth dates (which would hint that it is wrong person). In FTB there is warnings if you try but I dont know if it warns about this in the WEB tree.
I have no problem with the flagging as long as there ARE errors in the data. But what I have problems with is if a lot of correct data would be marked as "POSSIBLE DATA ERROR" because the reporter is the one with the error.
If a system with flagging data errors it must be a minimum to point out which data, correct data and the source that points this out. It should not be possible to point to a source that says "Because this SM I copied said so".
I cant say that it will be a problem with false reports but just the possibility of it makes me a bit doubtful. Perhaps if it was possible to keep some statistics if a reporter was leaving many false reports he would be at least informed by MH that he needs to check his own data before he report others and maybe as a ultimate "warning", restrict possibility to leave reports during a period with a message "Due to to many incorrect reports you are restricted to leave reports. You need to check your own data!". With to many incorrect reports I mean if the reporter leave incorrect reports to several trees and several persons desided by a algorithm made by MH. To make the report incorrect the reported need to leave a rejection of the report and need to tag it with a valid source. And now we have a circular problem: whatif you disagree about the sources... Which is the valid source? If I have the correct data I dont want it tagged!
I can only agree. It is to easy just to copy a smart match and do no check. People just seems to be happy to extend their tree without any check. The one with the most people in the tree wins. At least it seems so.
I almost never add a person from smart matches. I use the info to check out if I have missed something. When I can not find any proof that the info is correct I just leave it. I have 2 branshes in my tree I have copied in the beginning where I had contact with one of the tree mangers and we had much in common. I have not yet checked all but much in the beginning of the branshes so I at least is certain that we are related.
All my own research is checked with church records and I try to add all sources I can come over to follow each direct ancestor from birth to death.
I have also added some nobels that are related to me from literature and have the sources added to that literature. However I now and then get confirmations/rejections where the persons are clearly mismatched where older literature with outdated data gives my counterpart right. So how can you tell who is right when sources are missing. At least I know that I'm right :)
I would like to strike a blow for cleaning up personal web pages with the same people in the same pedigree.Too many genealogists forms copies/dual pedigrees,repetitions of the same data. One pedigree should be enough. If someone need copies put them in your computer. It is incredibly time-consuming to approve all the same match.
It gets confusing with all the same data. Please clean up- check out how many versions floating around of your family tree
Smar tMatch should be set to a highercomparison% total less than 45% should not be passed around.
Regards A frequent user o fMyHertage online version.
I have also voiced this concern on the product suggestions forum. About time MH did something. As a user, I would like to be able to set an option
(1) ignore trees that have not been updated in the last 1/2/5 or more years
(2) ignore sites with multiple trees, or at least ignore the older trees on sites with multiple trees, or even ignore the smaller trees on sites with multiple trees. Any of these, or better still, all of them, would fix the problem.
I agree that this is a very real problem. If it wasn't for the fact that I highly appreciate the links between MH and Geni I would have quit MH two years ago.
I routinely ignore sites that have more than one tree published but the problem is that doesn't work automatically on All of the trees form the site, just the one that as been physically ignored. I have one manager that has at least 12 trees and is continually creating others (presumably because he is producing versions centred on eachof his relatives). This wouldn't be so bad if once any profile has been accepted as a Smartmatch, that match would carry through to every tree that contained the same profile - ii is beyond logic as to why MH don't follow this process as it would clearly dramatically reduce the amount of of traffic. Could someone explain why every tree originalting from the same site doesn't automatically carry the result of a Smartmatch to all trees on that site that refer to that profile?
Turning to the related issue of abandoned trees - it is a feature of Geni that any tree that is not attended to for 18 months or so by the manager it is considered Abandoned and any other sharing Manager of a profile that has been abandoned by another manager can be claimed by that other manager. Why doesn't MH adopt the same rule? It wouldn't affect any existing profile and, in fact, would allow said profile be unpdated by managers that retain a working interest in that profile.
One final suggestion: Could we have check boxes down the left hand side of a page of Smartmatches? This would greatly speed up the process of going through a site with numerous Smartmatches which often contain a mixture of matches that deserve Confirming and others which Need to be Rejected. If there was a Check Box beside each prospective match, it would greatly simplify (and speed up) the confirm/reject process which at the moemnt is very time consuming because each individual Rejectio has to be done in isolation (you can only Confirm All or Reject All in batches - having Check Boxes would enable us to mark profiles for Rejection in a batch-wise process).
Ganske enig. Det er tidrøvende ud over alle grænser. En anden "synder" er: der anm. fx 15 matches, når den så er åbnet er der 25 og de sidste 7 - 10 er "falske/urigtige). Fx. at søskende matches forkert. Med hilsen JensL.
Denne gang skriver jeg på mit eget sprog, Dansk. Det er lettere, at udtrykke sig på sit eget sprog. Jeg håber, at nogle My Heritage folk fra Danmark ser dette, og gør noget ved det. Først vil jeg sige, at jeg har anvendt My Heritage siden 2009 online versionen. Jeg har været glad for det, men synes at det er blevet for besværligt med alle de fejl der er er kommet til. Det er min klare opfattelse, at My Heritage som udbyder har et ansvar for, at rydde op på server og i data. Det er ulideligt besværligt, at en bruger har tre, fire eller flere kopier af det samme træ. Der skal være en politik på My Heritage om, at der ikke online skal være kopier af de samme stam træer, og/eller anden form for multi arkivering online. Enhver der slægtsforsker, kan naturligvis gemme GEDCOM filer på sin egen computer i lige så mange kopier den enkelte mener at have brug for. Altså, smart match er udemærket, men ikke når det er uoverskueligt, eller fejl i visning, som sker når en forsker enten har slettet en person, eller ikke betalt, eller lavet andre fejl, som hele verden så skal deltage i at tjekke og kontrollere. Problemet er, at det er vanskeligt at skifte platform, fordi billeder ikke kommer med på GEDCOM filen, og så skal man indsætte alle foto en gang til ved opstart på en anden platform. Men det er den konsekvens der er. Hvis My Heritage ønsker høj vækst/service og mange kunder, skal der ske noget med data informationen, så det er til at have med at gøre.
Det kan gøres bedre på My Heritage. Ser frem til forbedringer på alle fronter.
Multiple trees with the same information are certainly a problem. I had regular matches from a person with sixteen trees!
In the case of one person with about six trees, his details are so consistent that I can check the first, glance at the second, and approve or reject all the rest based on that.
But your suggestion to set SmartMatches at a limit of 45%, in my own special case I would not want that; nearly all matches no matter how low the percentage are valid matches for my tree. I do sometimes get a run that are obviously not a match but not often.
Thank you all for your comments regarding this matter here on our forum.
We understand that this may be frustrating and even sometimes extremely.
MyHeritage provides the platform for users to create thier online family web sites and family trees and we do not keep track of how many trees duplicated or not that users have in thier family sites.
Having said that we do provide the option as already mentioned to "Ignore" the specific family tree that you are receiving the SmartMatches from and also the specific individual.
Other than that, users also have the option to contact the site manager of the matching family tree/site in order to communicate with them and point out thier error in duplicated family trees/individuals.
We hope that this informaiton has helped somewhat relieve a little of the frustration caused and again thank you for your comments as we do value these.
I recently contacted a 'no longer paid up' 'member' who had nearly every family member in a SmartMatch entered three times! The point is not that you can't contact them, SOMETIMES you can; but that they can, as it seems and as she believed, no longer make alterations to the tree.
I would hope that, at least in such unambiguous cases, MH would remove the duplicates if we ask. After all it is important to the company that people will join, and stay loyal; and accuracy of information is crucial to that.
Thanks for the infoaboutthis topic.My Heritage need a policy of observingthewebsitewhichare not paidorused.Theseshould be archivedon another serverso the datais not sentinappropriately tothe world.PeopletryMyHeritageand after a few test, many dont use it.MyHeritageshould havea policythat saysdeadline for the latestupdate.if a userhas not usedhis webfor more than 2years the user shouldbe accompanied by aconfirmation email.Naturally, Iagree thateveryone who hasa MyHeritagehasapersonal responsibility tomaintain order intheir own file and web.
I have contactedseveral usersand they saythat it is oftenbackproblem and the Famile Tree Builder inMyHeritage,and the many copiwebdoes notopen.In addition,it seems totallyout of control,that so manyjustpretendnothing.It's notOKto have towatch all thesmartmatchsthrough whenit can be easilysorted inMyHeritagesoftware control.As a user,I gettiredof the system.Best regardsElsebethVienberg
Vad menar du? SmartMatch har inget med "include files" att göta. SmartMatch ger dig bara upplysningar om, OM det finns en match mellan en person i ditt släktträd och samma person inågon annans träd. Kan du utveckla lite bättre vad du menar så skall vi se om vi inte får till det.