This research is about the basic methods in making computers understand human behaviour in an architectural space in regards to reaction to interaction between the machine and human. Its ultimate objective is to analyze the related technology making this series of works possible synthetically on the basis of information system within architectural territoriality. In the end it is expected to offer a theoretical basis to embody smart space, up-to-date and intelligent architectural space. There are two issues that motivate this research: what are the Housing Context and its Inference System, and how smart space can infer the Housing Context and react with proper response. The Husing Context consists of 1) state of user, 2) state of physical environment, 3) state of computational environment, 4) history of user-environment interaction and 5) architectural territoriality. Especially, spatial information of architectural territoriality is a significant key of HCIS. Spatial divisions and boundaries made of architectural elements or facilities determinate their own micro-territorialities. Ontologies are used to describe the Housing Context predicate. In this paper, we can say that the Housing Context and the Housing Log(history of user-environment interaction, a set of the Housing Context) written by ontologies can be a beneficial model of HCIS. Furthermore, we can develop the Housing Log Databases and its variable applications that have enabled to make simulating and analyzing tool of design, the Augmented Web Presence and the other helpful applications.