Back

---

Problem 1

Refer to Lab 1 (Problem 2), you have to incorporate business intelligence in your PVFC order processing system by adding a recommendation system.

Recommendation System

A recommendation system for an order processing system analyzes historical order data, user behavior, and item characteristics to predict and suggest products that a customer is likely to purchase next.

These systems

1. Improve customer experience
2. Increase Average Order Value (AOV) by identifying purchasing patterns and offering personalized recommendations.

Recommendation Techniques

Collaborative Filtering (CF):

Recommends items based on similar user behavior. If user A and B have similar order histories, the system suggests items purchased by B to user A.

Content-Based Filtering:

Suggests items similar to those a user has already purchased or viewed, based on item attributes (e.g., category, brand, or price).

Hybrid Systems:

Combines collaborative and content-based filtering to overcome limitations like the "cold start" problem (new users/items with no history).

Association Rule Mining

Identifies items frequently bought together, ideal for "frequently bought together" bundles or upselling complementary items.

Assumption: Sufficient numbers of orders are present in database.

Functionalities to be added

1. Display recommendation messages during Product Selection and Order Placement
   "Customers who bought this also bought..."
2. Reordering suggestions for customers
3. For employees/managers of PVFC forecasting demand for secondary items when a primary item's sales increase. (Inventory Management)