Ethical AI and AI Quality by Design - Building Responsible AI Systems

A conference talk from Data Science Conference Europe 2022 explores the intersection of ethical AI, emotional databases, and language analysis through the lens of measuring happiness across different languages. Dive into the development of emotionally annotated databases (EMODAT) and their applications in understanding human emotions, stress responses, and personnel evaluation. Learn about the positive bias in natural language through the analysis of frequently used words across multiple languages. Explore two primary research objectives: establishing minimal word sets for emotional state description in specific languages and developing cross-language happiness measurement rules. Gain insights into how big data analysis, machine learning, and automated data collection contribute to calculating happiness metrics through word correlation factors. Master key concepts including ethical AI challenges, DevOps principles, AI building blocks, machine learning methodology, responsible AI implementation, quality metrics, and software engineering practices. Understand how traditional software system metrics integrate with AI ethics standards and public intelligence principles to create more ethically sound AI systems.