The increasing use of smart devices allows us to extract massive streams of data, e.g., sensor streams, questionnaires, answers, annotations, etc. This information is crucial for the recognition of people's behaviours and habits. The main challenge is how to represent and organize such large scale, complex and heterogeneous data streams. This representation should allow for the management of all possible and unpredictable personal situations. The main goal of this paper is to propose a formalization of the personal situational context, showing how it can model real-life situations. The intuition is that, by collecting data from different people, we can populate the model and enhance the knowledge about those people by learning different aspects of their life habits. We start defining the abstract notions of the personal situational context and habits. Then, we provide an informal representation of such notions. Finally, we generate a universal ontological model of the situation context and habits, formally represented with an Entity Type Graph.