UML Diagram

This UML diagram illustrates the key components, relationships, and data flows within the Shopping Speed Bump system. It shows how different classes interact to provide a seamless intervention experience during online shopping.

Key components (hover to reveal details):

User Reflection Intervention

Browser extension, user preferences/goals, reflection prompts, and response tracking that create the intervention experience

Item & Wardrobe Management

Wardrobe items, shopping items, storage system, categories, and similarity engine for comparison algorithms

Tip: Use both horizontal and vertical scrolling to explore the full diagram. Use zoom controls to focus on specific class relationships and details.

100%

Math Challenges

The Shopping Speed Bump system relies on several mathematical algorithms and computational challenges to provide accurate, personalized interventions. A key innovation is that weights and scoring adapt over time - the system learns from each user's shopping behaviors, decisions, and stated goals to become more personally relevant and effective.

Similarity Scoring Algorithm

Challenge: Determining how similar a shopping item is to existing wardrobe items in a way that's personally relevant to each user.

Formula: Similarity = Σ(weight_i × match_i) where weights evolve based on user behavior

Threshold Optimization

Challenge: Setting the right similarity threshold to trigger interventions without causing alert fatigue.

Adaptive formula: New Threshold = Current Threshold ± (Dismiss Rate × Adjustment Factor)

Adaptive Weight Calibration

Challenge: Determining which attributes matter most for each individual user and adapting to their evolving preferences and goals.

Personalized learning: Initial weights → User behavior feedback → Goal alignment → Updated weights → Repeat