MusicRecommender

Workflow

Data Ingestion

1. user_ratings Table:
   • Populated by RatingDataParser, which parses user ratings in the specified file format and performs batch inserts into the database.
2. item_similarity Table:
   • Populated by DatabaseManager, which processes a similarity file to batch-insert records into the database.

Recommendation Generation

1. The RecommendationSystem queries user_ratings to retrieve user preferences.
2. It uses item_similarity to find items related to those the user has rated, applying a similarity threshold.
3. Recommendation scores are calculated based on a combination of user ratings and item similarity scores.

---

Database Overview

Tables

1. user_ratings Table

• Purpose: Stores individual user ratings for items, representing user preferences.

• Schema:

  • user_id (INT, PRIMARY KEY)
  • item_id (INT, PRIMARY KEY)
  • score (INT): The rating score assigned by the user.

• Usage:

  • Insertion: Handled by RatingDataParser.
  • Query: Provides user recommendations in conjunction with the item_similarity table.

---

2. item_similarity Table

• Purpose: Stores similarity scores between pairs of items.

• Schema:

  • item1 (INT, PRIMARY KEY)
  • item2 (INT, PRIMARY KEY)
  • similarity_score (DOUBLE): The calculated similarity score between item1 and item2.

• Features:

  • Add index to item1 for fast searching.
  • The similarity_score indicates the strength of the relationship between items.

• Usage:

  • Insertion: Managed by DatabaseManager.
  • Query: Used by the RecommendationSystem to find related items and compute recommendation scores.