Academic Marketplace Review Integrity and Fraud Detection Matrix
Systematize fake review identification, syndication anomalies, and tiered moderation actions across specialized scientific catalogs.
Ideal for Trust & Safety leads auditing suspicious review patterns, compensated testimonials, or bot-generated merchant ratings. It maps detection signals directly to automated and human escalation protocols.
Role: Trust & Safety Operations Architect and Marketplace Integrity Specialist with 12+ years designing behavioral fraud detection algorithms and moderation protocols.
Context
- Platform: {{marketplace_name}}
- Merchant Categories: {{vendor_categories}}
- Anomaly Signals: {{flagged_review_signals}}
- Response SLA: {{moderation_sla_hours}}
- Sanctions: {{enforcement_tiers}}
- Dispute Window: {{appeal_window_days}}
Task
Design a comprehensive review fraud classification matrix and moderation enforcement rubric that identifies, audits, and mitigates suspicious UGC manipulation on the marketplace.
Method
- Dissect the behavioral and technical vectors inside {{flagged_review_signals}} (e.g., IP clusters, velocity bursts, sentiment uniformity).
- Classify review fraud archetypes specific to {{vendor_categories}} (e.g., competitor sabotage, incentivized reviews, syndicated bot clusters).
- Define deterministic rule-based and machine-learning heuristic triggers for automatic review quarantine.
- Calibrate manual human review workflows within {{moderation_sla_hours}} based on vendor tier and listing revenue impact.
- Align detected severity levels with proportional penalties under {{enforcement_tiers}}.
- Establish an evidence preservation protocol to handle vendor challenges during {{appeal_window_days}}.
- Construct the final Fraud Triage & Enforcement Matrix detailing trigger criteria, automated safeguards, and escalation paths.
Constraints
- MUST establish zero-tolerance pathways for review rings and identity theft.
- MUST NOT recommend manual human review for low-severity signals that can be resolved algorithmically.
- False-positive mitigation mechanisms must be documented for every enforcement category.
- SLA adherence requirements must explicitly incorporate the constraints of {{moderation_sla_hours}}.
- Vendor appeal criteria must be clearly defined to ensure due process within {{appeal_window_days}}.
Output format
- Section 1: Fraud Pattern Taxonomy (150-200 words defining the top 3 marketplace threat vectors).
- Section 2: Review Integrity & Enforcement Matrix (Markdown table with columns: Fraud Pattern ID, Detection Signal, Risk Level, Automated Action, Human Review SLA, Vendor Sanction Tier, Evidence Standard Required).
- Section 3: Vendor Appeal and Remediation Lifecycle (Step-by-step audit protocol).
Self-review
- Are all signals from {{flagged_review_signals}} mapped to concrete enforcement actions?
- Are the punitive measures strictly aligned with {{enforcement_tiers}}?
- Is the workflow bounded by the operational window of {{moderation_sla_hours}}?
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.