Consumers face overwhelming choices when buying skincare, with thousands of products, complex ingredient lists, and trade-offs between price, quality, and skin compatibility. Many lack the tools to make confident, personalized decisions. Glowlytics is an intelligent decision-support system (IDSS) that helps users find the best fit skincare products by scoring each item across five key factors: quality, safety, value, popularity, and personalization. It continuously updates using web-scraped data and incorporates real user feedback to improve recommendations over time.
-
Clone the repository and navigate to the main folder
-
Run the following command from inside the ProductOptimization directory: python -m streamlit run app.py
-
Open the Streamlit app in your browser, input your skin type and concerns, and get personalized product recommendations with transparent score breakdowns
Five-Factor Scoring System: Each product is scored from 0 to 1 based on:
- Quality: Average user rating
- Safety: Ingredient risk
- Value: Price-performance ratio within product category
- Popularity: Brand reputation
- Personalization: Skin-type affinity from review analysis + concern keyword boost
Real-Time Feedback Integration: User ratings adjust future recommendations dynamically
Skin-Type Aware Recommendations: Review-based affinity scoring estimates how well products work for normal, oily, dry, combination, or sensitive skin
Up-to-Date Data: Nightly web scraping from Sephora ensures recommendations are based on the latest product listings, prices, reviews, and ingredients