This research aimed to design a mobile e-Learning forecasting system based on case studies using Multiple Intelligence analysis. The framework is divided into six modules. The first module describes a Patterns-based module. The second module describes the student Mapping module. The third module presents the Forecasting module. The fourth module presents the Learning portal module. The fifth module explains the Adaptive module. The last module presents the learning content based on the case study modules. In addition this paper introduces an example of an operational framework. The evaluation is comprised of two sections. The first section evaluates student achievement. The second section evaluates student prediction precision. The results of the first evaluation showed that students who studied via the mobile e-Learning forecasting system are more successful than students who studied via the normal e-Learning system with significance at 0.05 (Group 1, t = 7.577, p < 0.05; Group 2, t = 3.684, p < 0.05; Group 3, t = 15.190, p < 0.05). For the second section, the results considered the prediction precision percentage. We compared three algorithms: 1) J48 algorithm, 2) ID3 algorithm, and the 3) Naïve Bayes algorithm. The J48 algorithm had the highest value at 88.286%.