AI for Content Moderation at PayPal

Overview

Explore AI-driven content moderation strategies at PayPal, covering challenges, datasets, pre-trained models, and multi-language support for tackling hate speech and toxic content online.

Syllabus

• Introduction to AI in Content Moderation
Overview of Content Moderation at PayPal
Importance of AI in Handling Hate Speech and Toxic Content
Ethical Considerations in AI-driven Content Moderation
• Challenges in Content Moderation
Scale and Volume of User-generated Content
Accuracy Vs. Bias in AI Systems
Detection and Contextual Understanding of Diverse Content
• Datasets for Training AI Models
Overview of Available Datasets for Hate Speech Detection
Data Collection and Annotation Techniques
Addressing Dataset Bias
• Pre-trained Models and Algorithm Selection
Introduction to Pre-trained Language Models (e.g., BERT, GPT)
Model Adaptation and Fine-tuning for Content Moderation
Evaluation Metrics for AI Models in Moderation Tasks
• Multi-language Support
Challenges of Multi-language Content Moderation
Language Detection and Translation Techniques
Adapting AI Models for Multi-language Support
• AI Strategies for Detecting Hate Speech and Toxic Content
Text Analysis and Natural Language Processing Techniques
Leveraging Machine Learning Algorithms
Real-time Moderation and Delay-tolerant Approaches
• Implementation of AI Systems at PayPal
Architecture of Content Moderation Systems
Integration with Existing PayPal Infrastructure
Case Studies of AI in Action at PayPal
• Future Directions and Innovations
Emerging Technologies in Content Moderation
Trends and Future Research Areas
Continuous Improvement and Model Updates
• Course Conclusion
Summary and Key Takeaways
Discussion on Regulatory and Legal Perspectives
Final Q&A and Closing Remarks

Subjects

Conference Talks