Explaining data sources and features

Every AI prediction is based on data, making data sources a crucial part of explanations. 3 key aspects matter:

• Scope: overview of data collected
• Reach: whether personalized or aggregated
• Removal: whether users can delete or reset data

Consider a music app suggesting "Time to wind down" playlists at 9pm. Users might wonder how it knows their schedule. Explaining "Based on your typical listening patterns showing calmer music after 9pm" clarifies the data connection. Without this transparency, personalization feels creepy.

Data source explanations help users know system limits. If a fitness app only tracks steps and heart rate, users understand why it misses their weightlifting when calculating calories. This prevents over-trusting incomplete assessments.

Feature explanations show which data mattered most. A job matching AI might explain: "Match score based on skills overlap (45%), experience level (30%), location preference (25%)." This helps candidates understand why certain jobs ranked higher.

Privacy requires proper infrastructure beyond basic policies. Be transparent about data use while respecting boundaries.