How nilavra/mturk-crowdsourcing-pipeline works
Main Components
a) Client-Side Components (Frontend):
• Image Viewer Component (using iviewer)
• Annotation Interface
• jQuery-based UI Components
b) Server-Side Components:
• External HIT Handler
• Configuration Management
c) Database Layer:
• A database symbol labeled "Database System" (MySQL/SQLite)
• Connection to image list storage (referenced in sql_script)
d) MTurk Integration Layer:
• HIT Management
• Worker Management
• Results Processing
e) Processing Scripts:
• A container for "Processing Scripts" showing:
• Python Parser
• Shell Script Automation
Important Details
a) External Services:
• Amazon Mechanical Turk (as an external cloud service)
• Web Server (hosting environment)
b) File System Components:
• Image Storage
• Results Storage
• Configuration Files
Architectural Notes
• Notes about the technology stack:
• Frontend: jQuery, jQuery UI, Custom JS
• Backend: PHP
• Database: MySQL/SQLite
• Integration: MTurk CLI Tools
• Processing: Python, Shell Scripts
Additional Elements
• A security boundary showing public/private components
• Configuration management flow
• Worker interaction flow
• Data persistence paths