Documentation Practice

Effective documentation practices for artificial intelligence (AI) systems are crucial for transparency, accountability, and responsible use. Current research focuses on improving documentation standards for various AI applications, including clinical notes (using generative AI and NLP), third-party models and datasets (addressing inconsistencies in model cards and dataset cards), and cystoscopic media (emphasizing FAIR principles). This work aims to enhance the usability and reliability of AI systems by providing clearer guidelines and tools for documenting models, datasets, and their intended uses, ultimately improving both scientific rigor and practical applications.