Fuzzy Bayesian Context-Aware System to Reduce Electricity Consumption

2. Context And Context-Awareness

Context information characterizes the situation of an entity (Abowd et al., 1999). An entity can be a person, place, or object pertinent to the communication between a user and an application. Three different categories of contexts (computing context, network connectivity, communication bandwidth) are classified by Schilit, Adams and Want (1994) When a context provides relevant information to characterize the situation of an entity the system is context-aware

. For many researchers, context is primary, i.e., the main context, which can be the location(where), (who), time(when), activity(what), and secondary context derived from their relationships, constraints, and distances. The context-aware system makes the use of information about the state of the user and the environment and interprets and reacts to an individual's changing context. Context-aware systems help to solve the human-computer interaction problem in the internet of things and play a significant role in understanding sensor data (Abowd et al., 1999). IoT devices consist of a variety of sensors that gather information and send it back to actuators. The actuators decide with user attention and several assumptions. Context-awareness in IoT enables context information related to raw data to be stored so that interpretation can be made efficiently and intelligently with minimal human intervention according to the current situation (Perera et al., 2014). Improve the process to understand the context that significantly enhances user satisfaction by providing more customized and optimized solutions.

2 Context Life Cycle

The context-aware Internet of Things manages context information in four phases, as shown in Figure 1.

• 1.

  Context Acquisition: The context is acquired depending on the frequency, the source of context, the type of sensor (physical sensors, virtual sensors, or logical sensors) and the technique of acquisition (pull or push method).

• 2.

  Context Modelling: New context information is determined by context modeling in terms of attributes, characteristics, relationships with previously specified context, and quality of context attributes

Figure 1.

Context life cycle

• 3.

  Context Reasoning: A high level of contextual information derives from the raw context (unknown, ambiguous, imprecise, or erroneous). Different context reasoning techniques used to reason for the use of high-level context.

• 4.

  Dissemination: Dissemination of the inferred contexts to the consumer or to the application.