Since recommendation systems possess the advantage of adaptive recommendation, they have gradually been applied to e-learning systems to recommend subsequent learning content for learners. However, problems exist in current learning recommender systems available to students in that they are often general learning content and unable to offer personalized service. To overcome this, in the context of a learning style based on an Interpretive Structural Model (ISM), an adaptive learning path recommendation system is proposed comprising: (a) Fuzzy Delphi Method, (b) Fuzzy ISM; and (c) Kelly Repertory Grid Technology. The results show that the learning outcome with ALPRS is better than those from general learning course guided recommendation mechanisms, and the scores of system satisfaction with ALPRS and personal service are higher than 90%. Results of recall (95%), precision (68%), F1 index (45%) and MAE (8%) prove that ALPRS outperforms other approaches. Finally, three contributions are offered in this study: (1) a novel hybrid ALPRS is proposed and its practicability is tested; (2) a prototype gamification geometry-teaching material module is developed for the promotion in MSTE (Mathematics, Science and Technology Education) areas; and (3) the adaptive geometry-learning path diagram generated with ISM based on learning styles could offer a basis for further studies.