Authors: Dennis Mwanzia, Amos Kipkirui, Robert Mbau, Fiona Kungu, Maureen Kitanga, Edwin Muhia
MovieXplosion, a new streaming platform wants to improve their user satisfaction. The performance of the platform is dependent on how they can keep user engaged, one way to do this is by providing tailor-made recommendations to the users to drive them to spend more time on the platform. The project aims to develop a system that suggests movies to users. We will implement this using collaborative filtering, content-based filtering and hybrid approaches.
The current system that the platform employs does not provide suitable recommendations to users which has led to low user engagement, satisfaction and retention. The system also has no way of providing new users with good recommendations and the existing users do not receive tailor-made recommendations. The new system aims to bypass these issues and provide relevant recommendations to all users.
The data used was sourced from MovieLens, we used the small dataset due to limited computational power.
The data contains information about movies, ratings by users and other relevant information.
The data used can be found in the folder Data.
The data from the website had few issues with only 8 rows of missing values.
The dataframes from ratings.csv, movies.csv and tags.csv were merged.
F. Maxwell Harper and Joseph A. Konstan. 2015. The MovieLens Datasets: History and Context. ACM Transactions on Interactive Intelligent Systems (TiiS) 5, 4: 19:1–19:19. https://doi.org/10.1145/2827872
├── Data
├── Images
├──
├── notebook.pdf
├── presentation.pdf
├── README.MD
└── Movie Recommendation Systems.ipynb