Inspiration
Recommendation systems of amazon ,youtube etc.
What it does
You have to give a movie name and number of recommendations required .It finds the movies similar to given movie based on attributes like genre,actor ,director etc.After that it sorts the movies based on a similarity score and recommends required number of movies to the user.
How I built it
Firstly i scraped the movie data from IMDB site using a python script and stored it in a CSV file. Then i wrote java classes for calculating similarity ,sorting the list,modelling the data etc. This classes are:- ContentbasedRecommender.java ContentbasedDataModel.java ContentbasedSimilarity.java ContentbasedRecommender.java ContentbasedDemo.java (Main class)
Challenges I ran into
- During scraping some movie names were written in different languages like french ,German etc.So i had to face the encoding problems to print those movie names.
- Deciding which movie parameters to including and how much weight age to be given to each parameter in order to calculate a legit similarity score.
Accomplishments that I'm proud of
Scraping imdb site using python and beautifulsoup.
What I learned
I learned about how recommendation system works and how to scrap a site
What's next for Movie recommendation system
I want to make it personalized for every user so that user previous choices can also be used for recommendations.
Log in or sign up for Devpost to join the conversation.