A survey on interdisciplinary research of visualization and artificial intelligence

Abstract

With the breakthroughs in artificial intelligence (AI) technology, interdisciplinary research across AI and visualization (AI+VIS) has become one of the current research hotspots, providing inspiring theories, methods, and techniques for several core challenges in AI and big data analytics. On one hand, the innovative application of artificial intelligence technology has improved the efficiency of visualization, expanded the analysis capabilities, and provided powerful tools for big data visualization and analysis. On the other hand, visualization techniques enhance the explainability and interactivity of AI represented by deep learning, providing a reliable technical foundation for explainable AI. This paper introduces six important topics, including data quality improvement, explainable machine learning, intelligent feature extraction, automatic visualization layout and generation, intelligent interaction, and intelligent storytelling from two directions of "VIS for AI" and "AI for VIS", respectively. The research progress in the recent three years is analyzed. We also highlight the research trends of AI+VIS.