1

I am using Amazon Personalize (AWS Service) for making recommendation system. This service is expecting three datasets in the form of CSV from user named as

-> Interactions dataset
-> User dataset
-> Item dataset

I have imported User and Item dataset which is working fine but In the interaction dataset ,when I import it, it is giving me internal server error . I have three columns in my interactions dataset file, USER_ID , ITEM_D and TIMESTAMPS (In unix epoch format)

I don't know that why it is giving me internal server error ?

Hamza
  • 530
  • 5
  • 27

2 Answers2

1

In most of the cases, if the problem still exists after retrying, then raising the support ticket is the only solution.

Personalize support team should be able to help you, after you provide some details including ARN of Personalize dataset group.

It's common that Personalize is not very helpful with error messages and it's also nothing new that you can find a bug in the service itself. Thankfully the Personalize team usually fixes the issue quickly.

PatrykMilewski
  • 922
  • 8
  • 17
0

The timestamp column name must be TIMESTAMP and must of type long [1] in your interactions schema and CSV. It's also required. I suspect that since you're using TIMESTAMPS (plural) that Personalize is considering this an interaction metadata field/column and is complaining about a missing TIMESTAMP column.

[1] https://docs.aws.amazon.com/personalize/latest/dg/how-it-works-dataset-schema.html

James J
  • 621
  • 3
  • 6