A Dynamic Big Data Fusion and Knowledge Discovery Approach for Water Resources Based on Granular Computing and Three-Way Decision

Abstract

Abstract The purpose of this study was to achieve intelligent fusion and realize the unified modeling requirements of multi-source heterogeneous big data granulation and knowledge discovery in the field of water resources. This paper takes the management and decision-making data granulation and knowledge discovery driven by big data in the field of water resources as the research object, and uses the method of combining domain numerical simulation and model verification to systematically study decision-oriented big data multi-granularity granulation and knowledge discovery. The method reveals the mechanism and law of the transformation of management and decision-making paradigm driven by big data, and provides a complete solution method for the knowledge discovery of big data in various fields. The results obtained in this study include built a granulation mechanism and semantic fusion method of multi-source heterogeneous big data, and the multi-scale granular structure of big data is revealed, established and evaluated, and multi-granularity feature discovery and granulation method and multi-granularity uncertainty reasoning and knowledge discovery method. It was found that the formation mechanism of granular computing and three-way decision can be explained for dynamic big data fusion and knowledge discovery. The results indicated that the proposed dynamic big data fusion and knowledge discovery approach for water resources can reveal the semantic connotation and relationship of all kinds of resource objects in big data, so that to effectively support big data granulation and knowledge discovery in the field of water resources decision-making. Compared with the existing dynamic big data correlation analysis method, the proposed dynamic big data multi-granularity fusion method greatly reduces the data processing time, which fully shows that the proposed dynamic big data multi-granularity fusion and decision-making method has better performance.