A few months ago, I was writing about Market Basket Analysis that you can perform in Oracle Autonomous Database and Frequent Itemsets that can also be created with Oracle Analytics . When working with Oracle Analytics, one important part of Market Basket Analysis was missing - Association Rules and Recommendations . Oracle Analytics from version 6.0 supports this feature, therefore complete Market Basket Analysis can be done now. Let's take a look at the data flow we created for creation of Frequent Itemset. It consists of three steps: As usual, we begin a data flow by reading data. In our case we read database data table CARBO_TRANSACTIONS. This table has relatively simple structure: For our analysis only BASKET and UPC columns are important. In the second step, Database Analytics step with analytics operation Frequent Itemsets we define parameters for the algorithm. In the Outputs , we define output columns. In our case we are looking only for combinations of the two product
Oracle Analytics & Data Science