PETs Challenge: Repeated banks & accounts for different people


In the bank dataset, I find there are the same bank and account associated with different people, just like the following figure:

In this figure, the names are different, but the bank and account for these two names are the same. Is this a mistake or do I misuderstand the meaning of this dataset? I think different people should have their own bank name and account number. Can you help to clarify this?


Hi @hob,

Can you please clarify whether this is the dataset downloaded from the data download page, and how you are loading the data?

We do not see the rows you are showing with those account numbers. Instead, the rows with the same information that you are showing have account numbers 93698538821062294 and 93698538821062276, respectively.

Hi @jayqi ,

Thanks for your reply! I download the dataset from the page. I think there may be something wrong when unzipping or extracting the files, so the dataset is broken (changed). I download it again and the problem is solved. Thanks a lot for your help!

Hi @jayqi ,

I have found that not all the account information from SWIFT test set is provided in the bank dataset.

For example, there is an ordering account in SWIFT test dataset as GB3967905980802975863. However, there is no information about this account in bank dataset. The same situation also occurs to the account 11024064274693940, FR4471475542296975384 and so on.

The data set is downloaded from this page.

Hi @hob,

Accounts in the SWIFT transactions may not be present in the bank dataset. This is not an error.

1 Like

Hi @jayqi ,

OK. Thanks a lot for your clarifying!