Product description match for recommendation system

Hello, I am pretty new to knime and developing a workflow for a recommendation system with 2 tables a Buyers table and a Sellers table(datasets). I want a recommendation system where we can recommend the buyers as per the product description from the buyers’ dataset. The recommendation system works as if the buyers’ product description is somehow similar to the values of the seller’s description. Also, if we can show a similarity score between the buyer’s product and the seller’s descriptions, it would be great. I am attaching both the dataset’s tables. I have tried creating the model by changing the strings to document and using cosine similarity but that didn’t work. I would be grateful if anyone can help me in developing the model.
Buyer-Seller.knwf (9.7 KB)

Hi anjeesanjeet,

For a simple, basic proof of concept, you can start with the String Similarity node from the Palladian extension:


Thank you for the reply @qqilihq. I want to extract similar corresponding values from both tables. I want a model where I can input the buyer-required product and get a similar product from the seller dataset as per the description of the buyer-seller products.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.