Reputation System

Overview

The reputation system ensures trustworthy behavior is rewarded and spam/abuse is discouraged. Every user has a trust score (0-100) that affects how their submissions are processed and how much their reviews count.

Score Calculation

Trust Score = (approved_reports_ratio × 40)
            + (review_accuracy × 30)
            + (account_age_bonus × 15)
            + (verification_level × 15)

Factor	Weight	Description
Approved reports ratio	40%	% of submitted reports that were approved
Review accuracy	30%	How often their reviews matched final outcome
Account age bonus	15%	Longer history = more trust (max at 6 months)
Verification level	15%	Phone verified / ID verified / expert status

Score Tiers

Score Range	Tier	Benefits
0-20	New	Can submit, every report goes through full verification
21-50	Trusted	Reports skip expert review if AI confidence > 80%
51-80	Established	Reviews count 2x; can flag articles
81-100	Expert	Review count 3x; can become expert reviewer

Actions That Affect Score

Positive (+)

• Report approved: +10
• Review matches consensus: +5
• Flagging a fake report: +8
• Consistent activity over time: +2/week

Negative (-)

• Report rejected: -15
• Spam submission: -30
• Review against consensus: -3
• Flagged for abuse: -50 (or ban)

Database Schema

CREATE TABLE reputation (
  user_id TEXT PRIMARY KEY REFERENCES users(id),
  total_reports INTEGER DEFAULT 0,
  approved_reports INTEGER DEFAULT 0,
  rejected_reports INTEGER DEFAULT 0,
  reviews_given INTEGER DEFAULT 0,
  reviews_accurate INTEGER DEFAULT 0,
  trust_score REAL DEFAULT 0.0,
  last_activity_at TEXT
);