To implement INLI algorithm you will implement the following steps. 1. Read a batch of 100 tuples from TRANSACTIONS table as input data into a cursor. The Cursor is a user defined data type in PLSQL which works as a list and is used to store multiple records in memory for processing. 2. Read the cursor tuple by tuple and for each tuple retrieve the relevant tuples from PRODUCTS and CUSTOMERS tables of MD using PRODUCT_ID and CUSTOMER_ID as indexes on the both tables respectively and add the required attributes (mentioned in colour red in Figure 2) into the transaction tuple {in memory}. 3. The transaction tuple with new attributes is to be loaded into OW. Before loading the tuple into DW you will check whether the dimension tables already contain this information. If yes, then only update the fact table otherwise update the required dimension tables and the fact table. 4. Repeat steps 1 to 3 until you load all the data from TRANSACTIONS table to DW