/Movie-Recommender-and-Score-Prediction-System

Provides Movie Recommendations on the MovieLens ml-100k dataset using Collaborative Filtering

Primary LanguageTeXMIT LicenseMIT

Movie Recommender and Score Prediction System

Team Members

  • Akarshan Sarkar (akarshan2701)
  • Anuj Sharma (sharmaar12)
  • Muktinath Vishwakarma (foxx3)
  • Shikhar Sharma (kracwarlock)

Usage

All scripts are in 'Code 2' folder so switch to that directory

To find out rating of a movie for a user whose only the demographic information is known run this command:

demo(21,'M','student','94043','Copycat ',5)         % demo(age,gender,occupation,zipcode,movie_name,K_for_k-NN)

To run the whole code and compute everything for yourself open script.m and edit the dataset names in the following two lines:

dataset=read_dataset('../ml-100k/u3.base',0);
testset=read_dataset('../ml-100k/u3.test',0);

Now run this command:

script

To just check the predicted rating for a user using collaborative filtering and without using his/her demographic info, run these commands:

load('similarity.mat');
dataset=read_dataset('../ml-100k/u1.base',0);       % or u.data or any other file
rating=find_rating(1,10,dataset,similarity,10)      % rating=find_rating(userID,movieID,dataset,similarity,N_users_for_collab_filt);