Reference no: EM133139507

Java Collection Framework

1. Topic: Recommendation System

• What is recommendation System?
- Recommend content or products that users may be interested in by analyzing the user's existing behavior (viewing content, purchasing goods, etc.)
- Applications: OTT, e-commerce

• Two types of recommendation systems

- Content-based recommendation: By analyzing the characteristics of content that users have previously purchased, Recommend other content.

- CollaborativeFiltering:Finds other users B similar to user A's purchasing behavior, and recommends them to A based on B's purchases.

• Calculate similarity between two users in Collaborative Filtering

Jaccard use of similarity: |A∩B||A∪B|

Example:
. User A's set of purchases = {1 2 3 4 5}
. User B's set of purchases = {1 2 3 6 7 9}
.Similarity between A and B = 3 / 8
.Items that can be recommended to A = 6, 7, 9

• Input:
- File name, username to recommend, number of referring users, number of recommendations? v.txt x n m

• Composition of input file: User name Purchased items
• Suggested username: x ← String
• Number of reference user's n ← int
• Number of items to recommend m ← int

• Output

- Among the items purchased by n people with the highest similarity to x, items that x did not purchase are selected.
-m pieces are recommended, but recommended in order of frequency of purchase
- If there are several people with the same similarity, select them in ascending order by name.
-The frequency of purchase of recommended items should also be printed.
-If there are multiple items with the same purchase frequency, select them in order of their name.

Attachment:- Java Collection Framework.rar