Abstract: The evaluation technology of multi-source information fusion is an important method for the prediction of water inrush from the coal seam floor. However, this method has the deficiencies, such as high professional technical requirements, complicated and tedious evaluation process, and insufficient intelligence during its application. In this study, the design principle, design idea, development mode, and development method of a multi-source information comprehensive evaluation system for water inrush from coal seam bottom were systematically described according to the characteristics of the technical process of water inrush risk evaluation. The key functions of system were built, including the universal evaluation index system of water inrush from coal seam floor, the weight calculation methods and the risk evaluation model. Based on the GIS technology and the Visual Studio development platform, the third-party controls (WindowsForm control, DevExpress control, and ArcEngine control) were reasonably invoked, and a random forest machine learning method was introduced. In this way, a comprehensive evaluation system for the risk of water inrush from coal seam floor was developed, integrating multiple functions from data processing, storage, weighting operation, thematic map production to the visualization output of evaluation results. Thus, the "one" system was implemented from source data to visual presentation of evaluation results, which makes the evaluation technology of water inrush from coal seam floor more simple and efficient, and solves the problems of existing technology relying on multi-platform calculation, large separation of evaluation steps and low intelligent level.