Algorithmic Transparency Summary

1. Purpose of This Document

ReelCiety relies on algorithmic systems to organize, recommend, and distribute content in a way that balances relevance, safety, diversity, and platform integrity. This document exists to:

• Explain the role of algorithms in content visibility and discovery
• Clarify what data is and is not used in ranking decisions
• Demonstrate compliance with global transparency expectations
• Reduce misinformation about how ReelCiety operates
• Protect Nexa-Group from claims of opaque or unfair practices

2. High-Level Overview of ReelCiety Algorithms

ReelCiety uses automated systems to assist with:

• Feed ranking and ordering
• Explore and discovery recommendations
• Hashtag and topic surfacing
• Spam and abuse detection
• Visibility limitation for policy-risk content

These systems are designed to improve user experience and safety, not to guarantee reach, popularity, or monetization for any individual user.

3. Core Ranking Signals

ReelCiety’s ranking systems may consider a combination of the following non-exhaustive factors:

• User interactions (likes, saves, comments, shares)
• Watch time and completion rates
• Content freshness and relevance
• Topic and interest alignment
• Language and regional relevance
• Device and format optimization

Signals are weighted dynamically and may vary depending on surface (feed, explore, search) and user behavior.

4. Signals Explicitly Not Used

ReelCiety does not use the following categories as direct ranking inputs:

• Race, ethnicity, religion, or sexual orientation
• Political party affiliation
• Medical or health conditions
• Private messages or off-platform activity
• Financial status or creditworthiness

5. Safety and Integrity Layers

Algorithms include built-in safeguards to reduce the spread of harmful or borderline content. These layers may:

• Downrank content approaching policy limits
• Exclude content under review from recommendations
• Limit distribution of sensitive or graphic material
• Reduce reach for accounts with repeated violations

6. Visibility Limitation and De-Boosting

Not all enforcement actions involve content removal. In some cases, ReelCiety may apply visibility limitations that reduce algorithmic amplification without deleting content.

Visibility limitation may occur when content:

• Is sensational or misleading but not illegal
• Touches on sensitive events without reliable sourcing
• Approaches but does not cross policy thresholds
• Is posted by accounts with repeated borderline behavior

7. Human Oversight of Algorithmic Systems

ReelCiety does not rely solely on automated decision-making. Human teams:

• Audit algorithmic outcomes
• Review false positives and false negatives
• Adjust thresholds based on emerging risks
• Ensure alignment with safety and policy objectives

8. User Controls and Choice

Where available, users may influence their experience by:

• Following or unfollowing accounts
• Interacting with preferred content types
• Using mute, block, and report tools
• Resetting or adjusting personalization settings

9. Regulatory and Legal Alignment

ReelCiety’s algorithmic practices are designed to align with:

• EU Digital Services Act (DSA)
• UK Online Safety Act
• Emerging AI and platform transparency regulations
• Consumer protection and unfair practices laws

10. Limitations of Transparency

To protect users, prevent abuse, and safeguard intellectual property, ReelCiety does not disclose:

• Exact ranking weights or formulas
• Abuse detection thresholds
• Internal risk scores or trust metrics
• Security-sensitive system details

11. Updates and Evolution

Algorithmic systems evolve continuously. This summary may be updated as ReelCiety introduces new features, responds to regulatory developments, or improves safety mechanisms.

12. Contact

Transparency & Governance: transparency@reelciety.com
Compliance Office: compliance@nexa-group.org