RapidMiner

RapidMiner

Recommender Extension: Operator Apply Model

Contributor

Recommender Extension: Operator Apply Model

Hi,

I am using the RapidMiner version 5.3.013 with the Recommender Extension version 5.1.1
I'm trying to make some recommendations, so I followed some tutorials I found online.

To import the data to the process I used the operator Read Database. In the table, the attributes are defined as follows: user_id int(11) and item_id bigint(20).

In RapidMiner, after reading the data, the type of both is integer.
Then I used operators Set Role, User k-NN and Apply Model (Item Recommendation) like the tutorials exemplify.

My problem is that, after I run the process, the item_id that appears on results is different from that one that are in the database.
For example, the item_id in the database is 9876543210000 , but in the results of the process the item_id that appears is 2147483647 .
Anyone know why this happens? Did Apply Model (Item Recommendation) deals with bigint attributes?

Thanks for the time and help.
1 REPLY
Super Contributor

Re: Recommender Extension: Operator Apply Model

The integer type in RapidMiner does not support values above 2147483647. You should try to define the columns in RapidMiner as double. That should solve your problem.

Best regards,
Marius