KNIME Association Rule Learner (Borgelt) doesn't work

Hi,

I’m trying to create a Market Basket Analysis with data from books, where I would like to build a recommender system “if someone buys this book, they might also like this book”.

I’m trying to recreate this flow, but my data is not numbers, it is strings (book names):

I’m currently using the “Association Rule Learner (Borgelt)” - When I set the minimum set size to 1, minimum support to 10 and minimum rule confidence to 10 like in the flow i’m trying to recreate, I get the following error message:

ERROR Association Rule Learner (Borgelt) 0:5 External executable failed (error code 15): /Applications/KNIME 4.5.1.app/Contents/Eclipse/plugins/org.knime.ext.itemset_4.5.0.v202111151919/resources/executables/mac/x86_64/apriori - find frequent item sets with the apriori algorithm version 6.22 (2016.04.09) (c) 1996-2016 Christian Borgelt reading /private/var/folders/p5/krwfw9l9693b7lw34zky7z6r0000gn/T/knime_BA_122050/isInput8785497505707859250.txt … [65403 item(s), 197534 transaction(s)] done [0.06s].

Can anyone help what is might wrong? Cannot share the data since it is confidential.

Best

Hi,

Just to confirm, your cell splitter is producing a set as output correct?

And you have output that matches this structure:

Screen Shot 2022-05-03 at 9.11.37 AM

If yes to both, could you upload your workflow using a Table Creator node? (Please make a few fake samples that have the exact same structure as your original data). Also let me know your version of KNIME and what kind of OS you have.

Thanks

1 Like