Research on Intelligent Assessment Method for Low-Code Data Visualization Skills

Data visualization (DV) is a crucial technology for big data delivery and presentation, and an essential skill for data analysts (DAs) in the big data era. Effectively assessing DAs' visual operation skills has emerged as a significant research focus. Low-code DV skills enhance DAs' work efficiency and quality by offering a rich array of visualization components, yet skills assessment remains challenging due to its subjective and multidisciplinary nature. This study proposes an intelligent assessment method for evaluating lowcode DV skills using FineBI software, which stores the operation result in JSON format. To address the unique challenges posed by JSON paths containing “random codes” (system-generated unique identifiers, e.g., component IDs) and the acceptable variability in attribute values, we have devised an intelligent assessment method for DV skills. Experimental results indicate a 99 % accuracy rate in evaluating a set of DV skills test questions with reference answers. Our innovative low-code DV skills' intelligent assessment method offers a more efficient and objective method for DAs' skill certification, training, and talent selection.