Interest Adjustment Logic in the Evolution Mechanism of Social Stability Risks of Major Engineering Projects: A Study based on Grounded Theory

The design and realization of intelligent risk assessment system is an important topic for enterprise digital transformation. The study builds an assessment engine based on machine learning methods, and screens 35 key indicators through deep learning and integrated learning models, and the system assessment accuracy reaches 92.3%. The system developed based on microservice architecture supports high concurrency assessment request processing and is equipped with risk interpretation module to enhance credibility. Experiments show that the system significantly improves the efficiency of risk identification, shortens the assessment time from 4.2 hours to 1.5 hours, and provides an effective tool for enterprise risk management. The system shows good performance in practical applications and has important practical value.