Higher Education is facing disruptive innovation that requires, among other things, provision of a more effective and customized education service to individual students. In the field of Learning Analytics (LA), there has been much effort, in the form of the collection and thorough analysis of a variety of studentrelated datasets, to optimize learning performance and environments by means of personalization. The datasets include traditional questionnaire surveys, learning management system (LMS) log data of learner activities, and, more recently in the wake of the big-data-analytics trend, unstructured datasets such as SNS activities, text data, and other transactional data. Spatial data, however, is rarely considered as a key dataset, despite its high potential for characterization of students and prediction of their performances and conditions. In this context, the authors propose a new, spatial-data-driven student-characterization research framework. This vision paper describes spatial computing in its three, descriptive, predictive, and prescriptive modeling stages as well as its three challenges: (1) technical spatial data acquisition issues; (2) legal and administrative issues; (3) expansion of the application domains in which spatial data can contribute to improved modeling quality. With respect to each challenge, the on-going efforts are briefly introduced in order to substantiate the feasibility of the proposed research framework.