This paper introduces a novel approach to air combat intent recognition, emphasizing a model driven by the synergistic integration of expert knowledge and data. Departing from traditional knowledge base methods, our proposed framework leverages the interpretability of the Belief Rule Base (BRB) method and augments it with advancements in machine learning, particularly deep learning. The study explores the intricate interplay between BRB and neural networks, capitalizing on the strengths of each to create a robust decision support system. The belief rule base architecture is detailed, emphasizing fuzzy processing for handling information incompleteness. Additionally, the structure of the rule library, incorporating attribute importance, rule weights, and confidence levels, is outlined. Validation in the context of air combat target intention recognition demonstrates the model’s efficacy in combining expert knowledge and data patterns, presenting a compelling advancement in decision-making systems.