Skip to content

dxpr/ai_sorting

Repository files navigation

AI Sorting - Self-Learning A/B Testing for Content

Advanced A/B testing for Drupal Views that learns automatically. Instead of manually setting up A/B tests, AI Sorting continuously tests all content simultaneously using Thompson Sampling machine learning to surface winners while giving new content fair exposure.

Features

  • Advanced A/B Testing Engine - Powered by Thompson Sampling ML
    • Multi-variant testing of unlimited content pieces simultaneously
    • Zero setup required - no manual test configuration or statistical analysis
    • Continuous learning - tests never end, performance improves over time
    • Automatic statistical significance handling
  • Universal entity support - Works with nodes, users, taxonomy terms, media, custom entities, and external data
  • Cold start handling - New items get proper exploration scores automatically
  • Views integration - Simple sort plugin that works with any Views-compatible data source
  • Real-time learning - Continuous adaptation based on user interactions
  • Fail-hard debugging - No silent fallbacks, issues are immediately visible

Setup

  1. Install the module (requires RL module)
  2. Edit any View display (nodes, users, terms, media, custom entities)
  3. Add "AI Sorting" as a sort criteria
  4. Configure cache refresh rate and time window options
  5. Save - items immediately begin learning from user interactions

How It Works

  1. Track Engagement - JavaScript monitors when items are viewed and clicked
  2. Learn Patterns - Thompson Sampling identifies high-performing items
  3. Optimize Order - Best items automatically move to prominent positions
  4. Handle New Items - New content gets exploration scores for fair exposure
  5. Continuous Improvement - Performance gets better with every visitor

Beyond Traditional A/B Testing

Unlike traditional A/B testing tools, AI Sorting provides:

  • Multi-Variant Testing - Test unlimited content pieces simultaneously, not just A vs B
  • Zero Manual Work - No test setup, statistical analysis, or winner selection required
  • Continuous Optimization - Tests never end, performance improves over time
  • Smart Traffic Allocation - Reduces traffic waste by promoting winners faster
  • Cold Start Problem Solved - New content automatically gets fair exposure

Use Cases

  • Blog Posts - Surface most engaging articles first
  • Product Pages - Highlight items that convert best
  • News Feeds - Breaking stories get automatic priority
  • Resource Centers - Most valuable downloads rise to the top
  • Case Studies - Showcase most compelling success stories

Configuration

  • Cache Lifetime - How often content order refreshes
  • Automatic Cache Setup - Views cache automatically configured for optimal performance

Dependencies

Related Modules

  • RL module - Core Thompson Sampling algorithm and API for developers

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Packages

No packages published

Contributors 3

  •  
  •  
  •