01
Methodology Design
Contributed to the design of the recommendation methodology, including the selection of collaborative filtering and clustering approaches for analysing user preferences.
Data Science & AIAnime Recommendation System
BSc. Data Science Group Project · 2023
ICTlab, University of Science and Technology of Hanoi
Hanoi, Vietnam
Anime Recommendation System
A recommendation system using MyAnimeList ratings to generate personalised anime suggestions through collaborative filtering, matrix factorisation, and user clustering.
Recommendation Systems · Collaborative Filtering · K-Means Clustering
Project Overview
A course project exploring how collaborative filtering, matrix factorisation, and clustering can be used to generate personalised anime recommendations from anonymous user-rating data.
Role
Project Contributor
Methodology Design
Context
Artificial Intelligence
Course Project
Collaboration
10-Member
Student Team
Areas
Recommendation Systems · Data Analysis · Machine Learning
Tools & Technologies
Python · Collaborative Filtering · Matrix Factorisation · K-Means Clustering
My Contribution
02
Data Exploration
Participated in exploring the MyAnimeList dataset, examining anime metadata, anonymous user ratings, and the structure of user–item interactions.
03
Collaborative Development
Worked within a 10-member student team to organise the analytical workflow and document the recommendation process.
Project Workflow
The system transforms raw user-rating data into structured preference representations before generating personalised anime recommendations.
Step
01
Dataset Preparation
Prepare anime metadata and anonymous user-rating records from the MyAnimeList dataset.
Step
02
User–Item Matrix
Represent interactions between users and anime titles through a structured rating matrix.
Step
03
Preference Modelling
Use collaborative filtering and matrix factorisation to identify patterns in user preferences.
Step
04
User Clustering
Apply K-Means clustering to group users with similar rating behaviour and viewing interests.
Step
05
Recommendation Generation
Generate personalised anime suggestions based on related users, rating patterns, and cluster membership.