Life science product recommendation

Information: Product recommendation is rule based framework to suggests product to the customers who follow similar patterns. This helps business to reach specific group of customers rather than advertising it globally which will reduce the cost of marketing.

 

Requirement: Create a Databricks Pyspark code to build a product recommendation system by joining  d_product_revenue and health_insurance_claims tables using “Patient_ID” in health_insurance_claims and customer_id in d_product_revenue table

 

Note: cutomer_id in d_product_revenue table will be like GB002, GB001. Extract only the last digit and join it with health_insurance_claims table

 

Following are the rules for product recommendation:

 

Country-based*: Recommend the most common product (product_name) in the same country for customers who don’t have that product.

 

The final output should have two columns distinct “customer_id”, suggested “product_names“ (separated by commas).

 

Unity Catalog information: d_product_revenue and health_insurance_claims

 

Expected output: Databricks Pyspark code