Home      FAQ      Forum      Idea Exchange      Ask a Question      My Stuff      Help   
  
Prevent / Find / Merge record duplicates
There are a number of design questions we are considering with regard to this feature. We would love to get some more specific feedback on some of these questions:

* How would you want to specify the criteria for determining a duplicate?
* What action would you want to accomplish upon duplicate identification?
* Would you want the dupe to be merged automatically (similar as simultaneous record editing do), or just a message?
* Would that vary by how the record was entered (manual, web to record, API, etc.)?
ID
60
Category
User Experience
Author

Slava Shinderov  Staff 
Date Created
12/28/2006 6:51:05 AM
Date Updated
10/17/2017 12:47:08 AM
Status
New Idea
Score
60
Promoted By
Dale OliverPhilipp MatuschkaRick Cogley
ca@showtimegolf.comRainer HochkoepplerSlava Shinderov  Staff 
Comments
Slava Shinderov  Staff  5/18/2007 4:05:09 AM
Currently following trick could be used to detect possibly duplicate records:

* create the many-to-many relation to self;
* use Match Conditions to specify columns for duplicates criteria;
* add summary column to calculate # of records - when you have more than one you have a duplicate;

After that a view with filter on that column could be used to display a list of possibly duplicate records.
Rick Cogley 7/2/2010 12:48:53 AM
This would be wonderful to have. Re your questions:

* How would you want to specify the criteria for determining a duplicate?
=> Either full record, or, selected fields.

* What action would you want to accomplish upon duplicate identification?
=> Either take invasive action as in - Delete one, or merge them. Or, add a field like Duplicates and mark that field up with information so we can see what is a duplicate of what.

* Would you want the dupe to be merged automatically (similar as simultaneous record editing do), or just a message?
=> I would like the option to merge automatically, but I think it depends upon the record set, and, if I choose not to merge, I would like some field to show the info about the dupes. (by the way, the simultaneous editing warning is the cat's meow - very cool!)

* Would that vary by how the record was entered (manual, web to record, API, etc.)?
=> yes, I think it would have to vary by how the record was entered, if you mean to give a message right away. Non-interactive, it would be good to have some marker so the dupe could be seen and weeded out later.

Regards,
Rick
Rick Cogley 12/29/2010 8:08:29 AM
We are thinking about this lately, because one problem we have is, people enter contacts without first searching to see if the contact is already entered.

We are going to tighten up validation as Foresoft suggested to us separately, but it would be good to have some indication in the interface as you are entering, that there might be a dupe of what you are entering. The larger the db, the more likely this sort of thing gets, but also, it becomes more important.

One of the big challenges with this is, how do you define what is a dupe? You would have to have some way of specifying that, to the system.

Regards,
Rick


basenine 10/16/2017 10:07:46 PM
Hello Slava,

As it happens, without looking here first, I've replicated your steps to create a duplicates list.
I've done this because I wanted to change the Key Column for a new import of data that I'm about to process.
Trouble I was running into, was a warning that showed "Cannot convert to KeyColumn as Duplicates found"
Hence, I figured out a way (same as yours) to show the duplicates
I've gone through and deleted all duplicates until the view shows 0 records.

However, I'm still getting the same warning as before when I try to convert the KeyColumn
Any clues?
Best regards,
Brett

Slava Shinderov  Staff  10/17/2017 12:39:05 AM
@Brett please try to empty table recycle bin.
basenine 10/17/2017 12:47:08 AM
Of course! Been here before me thinks 😉
Thank you @slava
Feedback
 
Back to Search Results