wpForo v3 AI Edition Documentation

⌘K
  1. Home
  2. Docs
  3. wpForo v3 AI Edition Docu...
  4. AI Features
  5. AI Content Moderation
  6. AI Spam Detection

AI Spam Detection

AI Spam Detection

AI Spam Detection uses advanced machine learning to identify and handle spam content automatically. Unlike rule-based spam filters, AI understands context and can detect sophisticated spam that evades traditional filters.

AI Spam Detection Example

What AI Detects

  • Promotional spam – Advertisements, product promotions
  • Link spam – Excessive or suspicious links
  • SEO spam – Keyword stuffing, backlink attempts
  • Scam content – Phishing, fraud attempts
  • Bot content – Automated, templated spam
  • Off-topic promotion – Irrelevant self-promotion

Detection Levels

Level Score Range Meaning Typical Action
Clean 0-25% No spam indicators Publish normally
Uncertain 26-50% Some concerns Monitor or hold
Suspected 51-75% Likely spam Unapprove
Detected 76-100% Definite spam Unapprove + Ban

Configuration

AI Spam Detection Settings

Settings Explained

  • Enable AI Spam Detection – Master toggle
  • Detection Quality – AI processing quality
  • Use Forum Context – Include forum content for better detection
  • Minimum Indexed Topics – Required indexed content for context
  • Action on Spam Detected – What to do when spam is confirmed
  • Action on Spam Suspected – What to do when spam is likely
  • Action on Uncertain – What to do when unsure
  • Action on Clean – What to do when content is clean
  • Exempt Users with Posts – Minimum posts to bypass spam checks
  • Auto-ban Threshold – Unapproved posts before auto-ban

Moderation Report

Spam Detection Report in Moderation

View spam detection results in the moderation dashboard with:

  • Spam score and confidence
  • Specific spam indicators found
  • AI’s reasoning for the decision
  • Options to approve/reject

Best Practices

  • Enable forum context – Significantly improves accuracy
  • Use Balanced quality – Good accuracy at reasonable cost
  • Set exemption at 10-20 posts – Trusted users skip checks
  • Review false positives – Helps improve detection
  • Use auto-ban cautiously – Set threshold appropriately