https://github.com/dylandolevy/anime-globalization-fresh
Quantifying anime’s global boom (2010–2025) — MyAnimeList + Google Trends analysis
https://github.com/dylandolevy/anime-globalization-fresh
anime data-science kaggle notebook portfolio pytrends visualization
Last synced: about 1 month ago
JSON representation
Quantifying anime’s global boom (2010–2025) — MyAnimeList + Google Trends analysis
- Host: GitHub
- URL: https://github.com/dylandolevy/anime-globalization-fresh
- Owner: dylandolevy
- Created: 2025-11-20T02:52:09.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2025-11-20T04:02:55.000Z (8 months ago)
- Last Synced: 2025-11-20T05:25:31.109Z (8 months ago)
- Topics: anime, data-science, kaggle, notebook, portfolio, pytrends, visualization
- Language: Jupyter Notebook
- Homepage:
- Size: 302 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Anime Globalization — Data + Analysis (2010–2025)
**Short:** Analysis of how global interest in Japanese pop culture (especially anime) evolved from 2010–2025 using MyAnimeList metadata and Google Trends.
**Author:** Dylan Dole-Vy (GitHub: [@dylandolevy](https://github.com/dylandolevy))
**Status:** Working project — cleaned environment & notebooks (see `notebooks/`)
---
## Project Overview
This project builds a reproducible pipeline to:
- Clean and preprocess MyAnimeList metadata (`anime.csv`, `rating.csv`).
- Compute yearly metrics (titles produced, average score, membership).
- Pull Google Trends for *anime* (2010–2025) and aggregate to yearly.
- Merge and visualize global interest vs output, and compute correlations.
- Provide baseline predictive models for MAL score from metadata.
Notebook-driven, reproducible with a clean virtual environment.
---
## Abstract
Between 2010 and 2025, anime production and global audience interest expanded in tandem, reflecting Japan’s cultural export boom. Google Trends data show that worldwide searches for “anime” more than doubled after 2015, while annual anime releases rose by roughly 70 %. Correlation analysis (r ≈ 0.7) suggests that industry output tracks global demand, with a roughly one-year lag between peaks in search activity and production. Regional trends highlight surging growth in the U.S. and South Korea, confirming the globalization of anime beyond Japan’s domestic market. Genre analysis reveals a sustained dominance of Action and Fantasy, alongside a steady rise in Slice of Life titles—evidence that anime’s thematic diversity continues to broaden its international appeal.
## Reproduce Locally
1. Clone
git clone https://github.com/dylandolevy/anime-globalization-fresh.git \
cd anime-globalization-fresh
2. Create & Activate Environment
python3 -m venv .venv \
source .venv/bin/activate \
pip install -r requirements.txt
3. Jupyter Kernel
python -m ipykernel install --user --name=anime-fresh --display-name "Python (anime-fresh)"
4. Place Raw Data
Put anime.csv (and rating.csv if using) into data/raw/ (not committed if large). \
If you used Kaggle, keep kaggle.json at ~/.kaggle/kaggle.json.
5. Run the Notebooks
Open VS Code / Jupyter and run: \
notebooks/02_cleaning_exploration.ipynb \
notebooks/04_anime_modeling.ipynb \
notebooks/06_trend_analysis.ipynb \
The notebooks will save processed CSVs to data/processed/ and figures to outputs/figures/.
## Design Decisions & Notes
Raw large files are recommended to remain outside the repo. \
Use Git LFS for medium-large assets, or host raw data externally (Zenodo, Google Drive). \
Notebooks are the canonical analysis; the src/ folder contains reusable helpers. \
requirements.txt lists the package versions used for reproducibility.
## Key Findings (Summary)
Global search interest for “anime” rises sharply after 2015. \
Anime production (titles per year) increases in parallel. \
Strong positive correlation between production volume and search interest.
## Results Overview
Global Growth of Anime (2010–2025) \
Visualizing global search interest for anime (Google Trends) versus the number of anime produced each year. \
Interest and production volume both rise sharply after 2015, signaling the global “anime boom.”
Correlation Between Output, Scores, and Global Interest \
A correlation heatmap comparing yearly anime production, average scores, and worldwide search interest.
Anime Trends by Country \
(Full dataset and generation code in `notebooks/06_trend_analysis.ipynb`.)
🌐 Growth by Country (2010–2025)
Countries such as the US, KR, and GB show the largest percentage increase in anime search interest.
---
### 🎨 Genre Evolution
Trends in the five most common genres (2010–2025).
Shows how *Action* and *Fantasy* dominate globally, while *Slice of Life* gains steady momentum after 2018.
The US and South Korea show > 80 % growth in anime search interest since 2010, while Japan’s domestic interest remains stable — confirming globalization rather than domestic expansion. \
Production correlates ~0.7 with global interest, and the lag analysis suggests studios respond to rising demand with a 1-year delay. \
Genre analysis highlights the mainstreaming of Fantasy and Slice of Life themes, reflecting broader international appeal.
## License
This repository is distributed under the MIT License — see LICENSE (or change as desired).
## Citation & Data Sources
MyAnimeList dataset (Kaggle) \
Google Trends (pytrends) \
Jikan API (optional) for anime metadata enrichment