The Smart Mirror Ontology models the concepts, relationships and data involved in the functioning of an intelligent mirror system. It models users, sensors, products, medicine and environment information. It basic functioning includes:

• Turning lights on/off based on PIR presence detection
• Recognizing user activities and used products/medicines via RFID and camera
• Suggesting next steps (e.g., in the skin-care routines) based on detected activities or products
• Avoiding redundant suggestions by updating boolean properties (e.g., alreadyBrushed).
• Assessing skin condition through camera analysis and recommending suitable products.
• Monitoring product/medicine expiration dates via RFID and notifying the user
• The user can interact with the mirro using microphones

This ontology supports applications in smart health, smart homes and assistive technology.

The ontology presents 10 top-level classes, all directly under owl:Thing. Some examples of these classes are:

• Activities
• The bathroom
• The medicines the user takes
• The sensors the mirror exploits
• The user

Some classes present subclasses; for example, the Sensor class has as subclasses:

• The camera
• The PIR sensor
• The RFID reader
• The microphones

Also the Product class has some subclasses; particular highlight must be given to the subclass Cream, that has 3 subclasses itself which differ by the goal of the cream

The ontology defines 15 object properties, used to represent relationship between individuals. Some examples are:

• detectsPresence: PIR sensor -> User
• recognizedProduct: RFID reader -> Product
• takesMedicine: User -> Medicine

11 data properties are defined, which represent literal attributes. Some of them are:

• hasRoutineStepOrder, that specifies the step number of a product during the skincare routine
• hasTime, which specifies when a medicine must be taken
• hasDrynessLevel, that indicates the dryness percentage of the user's skin based on the camera analysis

The ontology counts 18 individuals and several object and data property assertion. Individual represent for example the user, Alice, some products, such as some creams, two medicines and the sensors. Property assertions describe the concrete relationship among the indivisuals.

The ontology is designed to include SWRL rules to support intelligent inferences. For example:

• A notification when the exiration date of a medicine or of a product is near
• A notification to hydratate the skin if the level of dryness is relevated too high
• The turning on/off mechanism for light

The SmartMirror Ontology aims to:

• enable automated reasoning on sensors and user data
• model interactions between the user and the mirror
• provide a foundation for ruke-based smart behaviours