(b) [4 points) Consider the following instance of a database containing transactions of items bought by customers at a grocery store. Each item in a transaction is given as a pair item_brand_name - item category, e.g., B, - 1. (There are 6 item categories, 11, 12, 13, 14, 15, and 16 and 8 brands, B1, B2, B3, B4, B5, B6, B, and B2) Customer_ID TID Items Bought 1 T1 {B.- 11, B2-12, B3 - 13, B4 - 1} 2 T2 {B4- 13, B3 - 12, B5 - 15, B6- 16, B,- 1} 1 T3 {B - 15, B3 - 12, B7-14, B6- 16} 3 T4 {B4-14, B2 - 12, B3 - 13} Let minimum support be 60% and minimum confidence be 50% and consider the following rule template: {item, item;2, ..., itemin] => {item;1, item2, item;} [s, c], where an item is one of {B,- 11, B2- 12, B3 - 13, B4-14, ...}, i.e., both the brand name and item category are taken into consideration, and s and c are the support and the confidence of the rule, respectively. List all frequent itemsets of size 3 at customer granularity level, i.e., the frequent itemset analysis should be done based on whether the items are bought by the same customer, not necessarily in the same transaction. Each items in the itemsets is one of {B4-11, B2 - 12, B3 - 13, B4 - 14, ...}. (You do not have to list the rules that can be inferred from these frequent itemsets.)