ONYEANUSI, CHISOM VICTOR (2026) DEVELOPMENT OF A FILM RECOMMENDATION SYSTEM USING MACHINE LEARNING ALGORITHM. Other thesis, Godfrey Okoye University, Enugu Nigeria.
![[img]](https://eprints.gouni.edu.ng/style/images/fileicons/text.png) |
Text
Film_Recommendation_Project_Final (3) (1)_102640.pdf Download (2MB) |
Abstract
The exponential proliferation of digital movie catalogues on streaming platforms such as Netflix, Amazon Prime Video, and ShowMax has created a severe information overload challenge for consumers, rendering manual film discovery inefficient and unsatisfying. This study presents the design, implementation, and evaluation of a Film Recommendation System using Machine Learning Algorithms, developed as a full-stack web application to deliver accurate, personalised, and explainable movie suggestions to registered users. The system integrates three complementary recommendation techniques: Singular Value Decomposition (SVD) for collaborative filtering, Term Frequency-Inverse Document Frequency (TF-IDF) cosine similarity for content-based filtering, and a dynamically weighted hybrid model that combines both approaches to mitigate the cold-start problem prevalent in sparse rating environments. The MovieLens 100K benchmark dataset, comprising 100,000 ratings from 943 users across 1,682 movies, was employed for model training and offline evaluation. The backend was implemented in Python using the FastAPI framework and the Surprise library for SVD modelling; the frontend was built using React.js with Tailwind CSS; and data was persisted in a PostgreSQL/SQLite relational database. User acceptance testing with 15 participants produced a mean usability score of 4.2 out of 5.0 and a recommendation relevance score of 3.8 out of 5.0. The study confirms that a weighted hybrid approach combining matrix factorisation with content-based metadata modelling yields superior recommendations compared to either method independently, and that explanation features significantly enhance user trust and system perceived quality. Keywords: Recommendation System, Collaborative Filtering, Content-Based Filtering, SVD, TF- IDF, Hybrid Model, MovieLens, FastAPI, React.js, Machine Learning.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Computing And Information Technology (FACIT) |
| Depositing User: | Cynthia Ugwuoti |
| Date Deposited: | 24 Jul 2026 14:22 |
| Last Modified: | 24 Jul 2026 14:22 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6004 |
Actions (login required)
 |
View Item |
