Discount Quality for Responsible Data Science : Human-in-the-Loop for Quality Data

mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm; mso-border-alt: none windowtext 0cm;

This open access book presents a “discount”-quality approach to data preparation for scientific data analysis, a critical prerequisite in modern applications such as data analysis projects, Artificial Intelligence, and Machine Learning. It discusses advanced techniques for fostering responsible data science, i.e., designing sustainable data analysis pipelines based on Human-In-The-Loop (HITL) approaches to achieve high-quality data. It investigates developing task- and context-driven sustainable approaches for data preparation, drawing on methods and theories to reduce annotations and processing space and time, and considering the estimation of the necessary human computing effort and the requirements of the specific task. The contributions address key aspects related to data ecosystems, data preparation pipelines, data quality evaluation and improvement, on-demand approaches to data preparation, data enrichment, and human factors in data preparation.