The use of context is already a reality in different application fields. One such example is the Robotics field, where sensors are used to capture information about the surrounding environment, allowing the system to react to changes. But also in the multimedia communications field, some work has recently emerged to allow combining sensed context and use that combined information to decide how to adapt the content. This paper describes the MULTICAO ontology, which lies at the core of an Adaptation Decision Engine, providing it with a formal description of the surrounding environment in a similar way as humans would describe it. This way the system has higher chances to take adaptation decisions that best meet users' expectations.