In recent years, there has been a dramatic proliferation of research concerned with the ubiquitous learning (u-learning). The u-learning systems have to be continuously evaluated and improved for ensuring the system reliability. Therefore, this work based on meaningful learning aspect to propose a u-learning evaluation model. The model blends features of u-learning and meaningful learning to construct a hierarchy decision model. According to the hierarchy structure, domain experts can develop AHP-based questionnaire survey to collect learners' opinions. Following that, system developers can realize the relative strength and weakness of the u-learning system from a meaningful learning viewpoint by analyzing the surveyed data, and they can further to improve and refine current u-learning systems accordingly. Consequently, existing u-learning systems can be revalidated by our evaluation model, and then based on the produced suggestions to improve toward the meaningful learning.