Abstract
To address the critical challenges of decision-making delays and accuracy deficiencies faced by tourism enterprises in dynamic environments, this study develops an intelligent decision support platform based on multi-source heterogeneous data fusion. The research proposes a novel hybrid data fusion algorithm that integrates weighted averaging, Bayesian inference, and Dempster-Shafer evidence theory, establishing a comprehensive framework for processing three types of heterogeneous data from diverse sources. The main theoretical contributions include: (1) a hybrid data fusion algorithm achieving superior accuracy through adaptive weight adjustment mechanisms, (2) a multi-dimensional environmental adaptability assessment framework tailored for tourism enterprises, and (3) real-time decision optimization models with self-learning capabilities. Empirical validation across three different types of tourism enterprises demonstrates significant improvements in decision accuracy (18.96% average increase), response time reduction (78.4%), and environmental adaptability enhancement across market responsiveness, risk prediction, and resource optimization dimensions. The results indicate substantial cost-benefit ratios (3.4:1) and sustained competitive advantages, validating the platform's effectiveness for tourism enterprise digital transformation and strategic decision-making in volatile business environments.