1. A Comparative Study on Ethics Guidelines for Artificial Intelligence Across Nations
- Author
-
Shiang Yao Liu, Li Yun Chang, Yin Ling Wei, and Tony Szu Hsien Lee
- Subjects
Scheme (programming language) ,Knowledge management ,Computer science ,business.industry ,Educational technology ,Term (time) ,Data visualization ,Content analysis ,Similarity (psychology) ,Key (cryptography) ,Tag cloud ,business ,computer ,computer.programming_language - Abstract
This study aimed to investigate the commonality and differences among AI research and development (R&D) guidelines across nations. Content analysis was conducted on AI R&D guidelines issued by more economically developed countries because they may guide the trend of AI-based applications in education. Specifically, this study consisted of three phases: 1) information retrieval, (2) key term extraction, and (3) data visualization. First, Fisher’s exact test was employed to ensure that different AI R&D guidelines (e.g., the latest ones in the US, EU, Japan, Mainland, and Taiwan) were comparable. Second, the Key Word Extraction System was developed to retrieve essential information in the guidelines. Third, data visualization techniques were performed on key terms across multiple guidelines. A word cloud revealed the similarity among guidelines (e.g., key terms that these guidelines share in common) while a color-coding scheme showed the differences (e.g., occurrence of a key term across guidelines and its frequency within a guideline). Importantly, three key terms, namely, AI, human, and development, are identified as essential commonality across guidelines. As for key terms that only extracted from particular guidelines, interestingly, results with the color-coding scheme suggested that these key terms were weighted differently depends on the developmental emphasis of a nation. Collectively, we discussed how these findings concerning ethics guidelines may shed light on AI research and development to educational technology.
- Published
- 2020