How does the Similarity score is being calculated in Match Transform?

[former_member690658]

While calculating Match score based on Weight Method. To verify the calculation I am trying manual calculation, But for that 'Similarity Score' needs to be calculated first.

For the records where number of characters are same it is quite easy to get the 'Similarity Score'. For example

Record A Record B Similarity Smith Smith 100% Smith Smitt 80% ms@ sap.com msmith@ sap.com 84% ??

https://help.sap.com/viewer/e54136ab6a4a43e6a370265bf0a2d744/4.2.14/en-US/57353eeb6d6d1014b3fc9283b0...

How in Email address similarity is coming as 84%.? Please help me with the calculation behind it.

[Julian_Riegel · Accepted solution]

Hi there,

this is nothing you can check by yourself. DS uses internal code/algorithms to retrieve such values.

Nevertheless, the string similarity most likely is calculated by a variation of the basic of Smith Waterman Algorithm.

If you are really keen on testing its basic functionality you could visit this site and see which DS output gets the closest:

https://asecuritysite.com/forensics/simstring

Regards,

Julian

[former_member690658 · Accepted solution]

Thanks @Julian Riegel

I was expecting the same after going through SAP guides. Just wanted to confirm that my understanding is correct about it.