PratikhyaManas / Movie-Recommender-System

In this project, a movie recommendation system is built using Item-based Collaborative Filter.

Movie-Recommender-System

1. Recommender systems are algorithms designed to help users discover movies, products, and songs by predicting the user’s rating of each item and displaying similar items that they might rate high as well.

2. The objective is to show customers content that they would like best based on their historical activity.

RECOMMENDER SYSTEMS: USER-BASED COLLABORATIVE FILTER

1. User-based collaborative filtering works by building a matrix of every piece of content that users bought or viewed.

2. Similarity scores are then calculated between users to find similar users to each others.

3. For similar users, content that have not been viewed or bought are recommended to users that haven’t seen them before.

RECOMMENDER SYSTEMS: ITEM-BASED COLLABORATIVE FILTERING

1. Based on User #1 and #2, they both watched and liked Titanic and a walk to remember.

2. Item-based collaborative filtering will correlate both movies together based on user #1 and #2 behaviour.

3. User #3 watched “Titanic” and did not watch a “Walk to remember”, so the recommender system will recommend it for him/her.

PROBLEM STATEMENT

1. This notebook implements a movie recommender system.
2. Recommender systems are used to suggest movies or songs to users based on their interest or usage history.
3. For example, Netflix recommends movies to watch based on the previous movies you've watched.
4. In this example, we will use Item-based Collaborative Filter