How We Scraped and Analyzed the Global Top 100 Most Beautiful Women with Python

We explore a Python‑based workflow that extracts, cleans, visualizes, and evaluates the "Global Top 100 Most Beautiful Women" ranking published by an overseas media outlet.

Evaluation Criteria

The ranking score is calculated by the formula:

Total Score = 0.3 × fan votes + 0.5 × official facial proportion score + 0.2 × (personality, body, charity, etc.)

Data Acquisition

Because the page is static, we can directly parse its HTML source. The script uses requests to fetch the page and BeautifulSoup to locate each li element whose id varies with the ranking position. A loop constructs the appropriate id values, extracts the name, nationality, and photo URL, and saves the images locally with a self.downloadImg helper.

Data Analysis

1) Continent Statistics

We aggregate the data by continent, counting how many entries originate from each region. Countries with dual nationality (e.g., "Filipina‑American") are counted twice.

2) Nationality Statistics

We normalize nationality names (e.g., converting "Thai" to "Thailand") to match the standard country list required by pyecharts. After cleaning, the visualization shows the United States leading the count, followed by the United Kingdom, with China and South Korea appearing without dual citizenship.

3) Profession Analysis

The list contains only three professions: model, actor, and singer. Actors dominate because facial appearance carries the highest weight in the scoring formula.

4) AI Beauty Scoring

We employ Baidu’s face‑recognition API. The FaceScore function reads each downloaded image as binary, encodes it with base64, sends it as a POST request, and extracts the returned beauty score.

The top‑5 scores are displayed in the following chart:

Overall, the tutorial shows how to combine web scraping, data cleaning, visual analytics, and AI‑driven scoring to derive insights from a seemingly entertainment‑focused ranking.