The ability of discovering frequent behaviours of the users allows an environment to act intelligently, for example automating some devices’ activation. Moreover, such frequent behaviours could be used to understand and detect bad or unhealthy habits. Such a discovering process must be as unobtrusive and transparent as possible. In that sense, the ability of inferring interesting information from sensors installed in the environment plays an essential role in order to provide the discovering process with meaningful data. The importance of this system is clear due to the fact the process of discovering frequent behaviours will totally depend upon the actions/activities identified by such a system. This development reinforces the link between context-awareness and human behaviour understanding as it can perceive a current situation, compare it to typical behaviour, and differentiate between the two.
TopIntroduction
Intelligent Environments (IEs) are “…digital environments that proactively, but sensibly, assist people in their daily lives” (Augusto et al. 2007). They offer an opportunity to blend a diversity of disciplines, from the more technical to those more human oriented, which at this point in history can be combined to help people in different environments, at home, in a car, in the classroom, shopping centre, etc.
One of the hidden and most important assumptions in IEs is that they are pioneering a transition from techno-centered systems to human-centered systems. IEs suppose a change of roles in the relationships between humans and technology. Unlike current computing systems where the users have to learn how to use the technology, an IE adapts its behaviour to the users, even anticipating their needs, preferences or habits.
For that shift to take place, an environment should learn how to react to the actions and needs of the users, and this goal should be achieved in an unobtrusive and transparent way. In order to provide personalized and adapted services, the requirement of knowing the preferences and frequent habits of users is clear. Thus, the ability to learn patterns of behaviour becomes an essential aspect for the successful implementation of IEs, because knowing such patterns allows the environment to act intelligently and proactively. In IEs, learning means that the environment has to gain knowledge about the preferences, needs and habits of the user in order to better assist the user (Galushka et al. 2006; Leake et al. 2006).
IEs also assume that the process of acting intelligently over the user cannot disturb them. Otherwise, if a user is continuously disturbed, it could be a reason to reject the system (Liao et al. 2004; Pollack 2005). This requirement demands IEs to be conscious of the current situation in each moment (Ramos et al. 2006). Context aware environments are concerned with the unobtrusive acquisition of context (e.g., using sensors to perceive different situations), understanding of context (e.g., inferring the actions/activities the users are doing), and making decisions based on the recognized situation (e.g., automating the activation of a device).
Being aware of the importance of these two areas, it must be said that they are not independent areas but they must be combined to achieve a real intelligent environment. Taking the learning system as the base to provide the environment with intelligence, this chapter analyses the influence of context awareness in the learning process. For that, using a real scenario, we identify the steps of the learning process where it is necessary to apply context awareness techniques to allow the environment to provide personalized and adapted services at the right time. Let us consider a scenario that illustrates an IE that makes the life of the users easier and safer.
Michael is a 60-year-old man who lives alone and enjoys an assistance system that makes his daily life easier. On weekdays, Michael’s alarm goes off a few minutes after 08:00 a.m.; approximately 10-15 minutes later, he usually steps into the bathroom. At that moment, the lights are turned on automatically. On Tuesdays, Thursdays and Fridays, he usually takes a shower; Michael prefers the temperature of the water to be around 24-26 degrees Celsius in the winter and around 21- 23 degrees Celsius in the summer. When he finishes taking a shower, the fan of the bathroom is turned on if the relative humidity level of the bathroom is high (in Michael’s case >70%). Before he leaves the bathroom he turns off the fan and the lights. When he goes into the kitchen the radio turns on so that he can listen to the news while he prepares his breakfast. When he is preparing his breakfast the system reminds him that he has medicine to take. He leaves the house 15-20 minutes after having breakfast. At that moment, all the lights are turned off, and safety checks are performed in order to detect potentially hazardous situations in his absence (e.g., checking if the stove is turned on), and if needed, the house acts accordingly (e.g., turning the stove off).
The remainder of this chapter is organized as follows. Section 2 provides a literature review of related works. Section 3 introduces the architecture of the system that discovers frequent behaviours. Then, Section 4, 5, and 6, respectively define the different layers of such a system. Finally, conclusions and future research directions are identified.