Abstract: Aiming at the problems of scattered information, delayed supervision, and difficulty in risk prediction in current house expropriation management, an intelligent scheduling platform integrating multi-source data is designed to improve the refinement and collaboration level of expropriation management. The platform adopts a microservice architecture, integrates WebGIS technology, relational databases, and their spatial extension functions, uniformly organizes and manages basic geographic information, expropriation thematic data, structured business data, and unstructured materials, and constructs an expropriation dynamic early warning model combining spatial analysis and business rules. Verified through the expropriation work in Hefei, the platform realizes the "one-map" visualization and dynamic update of expropriation information, and the data collection efficiency is improved by about 70%; the dynamic early warning model assists managers in identifying potential risk points in advance, and the average decision-making response time is shortened by about 50%. This platform effectively breaks down data barriers, improves the transparency and decision-making efficiency of expropriation management, and provides a feasible technical path for urban renewal governance.