Hello All, I ended up here after looking at : https://www.zingg.ai/documentation-article/identifying-duplicates-in-snowflake-with-zingg So the entity resolution works on just one input table? What if we had multiple different input tables? How would the entity resolution needs to be defined? Example: We have Customer table and then we have Reviews table, A purchase transaction table. What if we want to unfiy across all these three domains?
If all three have same schema you could give multiple input pipes
If they have different schemas => they are 3 different entity resolution problem, so you will have to train 3 different models each catering to one particular table/schema
alternatively you can join them together into 1 table and than use that to train the model
Thank you for the response. Think of an Retail company scenario. Customer domain, Sale transaction domain and Product Reviews domain. They wont have the same schema but all of them will have some set of PII data. But the idea is to implement identity resolution to be able stitch across. Is something like this done via ZIngg? Joining them together into one table doesnt feel like a scalable solution.
Zingg Community needs denormalised datasets to resolve entities effectively. On the Zingg enterprise product, we provide the ability to bring in normalised data with their pii through a phase called relate. Happy to chat if that’s of interest!
Thanks for the clarification. Can i read more of the difference between community and enterprise editions? Please share