A Study of Interactive Decision-Making Techniques in Multimodal Big Data Visual Analytics

Abstract

With the rapid advancement of big data technology, processing and analyzing multimodal, heterogeneous data have become a central challenge in the field of decision science. Multimodal big data visual analytics aims to transform complex data into actionable decision insights by integrating visual interaction with computational models. This study focuses on interactive decision-making techniques in this domain, systematically exploring their theoretical foundations and technical architecture. The research first analyzes the core challenges associated with the representation learning and heterogeneous fusion of multimodal data, the cognitive theoretical basis of visual analytics, and the construction of decision task-driven analytical models. Subsequently, it constructs a technical framework for visual analytics oriented toward interactive decision-making, encompassing key technologies such as multi-scale dynamic visualization, interactive feature exploration, uncertainty quantification, and decision reasoning. Finally, it proposes design principles for decision-maker-centered system integration, methods for enhancing interactive intelligence, and a multidimensional performance evaluation framework. This study provides theoretical references and technical pathways for building high-performance interactive decision support systems.