To address the challenges of non-uniform fracture propagation，low fracture prediction accuracy，and low cumulative oil production in shale oil well factory-mode fracturing，dimensionality reduction，entropy，and grey relational analysis methods were applied to establish a quantitative evaluation methodology for geo-engineering dual sweet spots. A competitive propagation mechanism for multiple fractures in well factory mode was proposed. The Monte Carlo method was applied to stochastically sample parameters such as staging/clustering，fracturing sequence，perforation design，and pumping schedule. Coupled with dual sweet-spot characterization and four-dimensional in-situ stress，a multi-parameter coupling approach was used to compute fracture parameters. The finite volume method was applied to simulate and predict various development indicators，and the optimal combination of fracturing parameters was selected. The results indicate that the geo-engineering dual sweet-spot value for high-quality reservoirs is greater than 0.79；the four-dimensional in-situ stress field reveals stress shadowing effects among fractures，clusters，stages，and wells；microseismic monitoring confirms that the numerical model based on the competitive multi-fracture propagation mechanism predicts fracture parameters with an average relative error of 9.3%，which is 10.2 percentage points lower than that of commercial software on average；the staged cumulative oil production from the multi-parameter coupled unconventional fracturing is on average 13.0 percentage points higher than that from conventional fracturing，demonstrating favorable application performance.