[00:00:01] Hello. In today's video, I want to show you a new feature in MOSTLY AI, the entity relationship diagram for multi-table synthetic data setups.
[00:00:11] As a basis for this video, I'm using a tutorial, which we have published on our website. It shows you how to create synthetic data from a multi-table setup using the Berka data set.
[00:00:23] The Berka data set is relating to bank account transactions. Here, you have the account and then transactions. You also have loans and orders.
[00:00:35] Because there's a many-to-many relationship between accounts and clients, there's a disposition table in the middle that takes care of that. Then we use the district to make sure that the correlation between client and account is preserved as much as possible.
[00:00:56] I have already loaded these tables into my job, and I just want to show you the relationship diagram here.
[00:01:08] You can very clearly see how the account is the parent table for transaction, order and loan, how there is this disposition table here that links the client to the account because of the many-to-many relationship, and then the cards are also depending on the disposition.
[00:01:27] With this feature, it makes it a lot easier for you to understand how the tables relate to each other, and then create synthetic data with referential integrity between all of them. Thanks a lot.