What is Context Aware Computing ?

1.0 Context Aware Computing : 




Ubiquitous computing  [ Weiser 1988] is a model of human- computer interaction , in which information processing has been thoroughly integrated into everyday objects and activities. It is the method of enhancing computer use by making any computers available throughout the physical environment, but making them effectively invisible to the user. One of the distinguishing features of Ubiquitous Computing is that the computation is a part and parcel of everyday life. The computing services required for user depends on user’s  social and personal context. Thus the ubiquitous computing systems can provide more meaningful and useful services provided the systems are context aware in nature.  Ubiquitous Systems are a store house of sensors and devices without context aware computing. It is the context aware methodologies which 

•         makes ubiquitous systems aware of situations of interest
•         enhances services to users
•         automates systems and reduces obtrusiveness
•         customizes and personalizes applications.

Albrecht Schmidt in his report[1] highlights the importance of context aware computing as: “Context is essential for building usable Ubiquitous Computing systems that respond in a way that is anticipated by the user”.

1.1 Definition of Context: Definitions given by earlier works and standard dictionaries agree on the key idea that contexts describe situations. This definition clearly states that context is always bound to an entity. The entity itself is regarded as something that is relevant to the interaction between a user and an application. The user-application relationship is rooted in the traditional notion of an application, but not limited to it.

Dey and Abowd (2000) have also confirmed this by defining context as: ‘‘Any   information that can be used to characterize the situation of an entity. An entity is a person, a place, or a physical or computational object that is considered relevant to the interaction between a user and an application, including the user and application themselves.’’ [2]

Pascoe defines context as “Context could be generally described as the subset of physical and conceptual states of interest to a particular entity.” Contextual information is related to a certain entity.[3]

Context in this report is represented as a set of attributes which characterizes the situation of an entity.

For Example the context of user can be represented as:

User_Context = {User_Name, User_Role, User_Age, User_Interest, User_Mood, User_PhysicalPosture, User_PhysicalHealth, User_ Activity, USer_Location, Date, Day, Time}.

A set of attributes are said to describe a context if and only if they are capable to answer the questions mentioned below:

1. Who is he?            (Identity)

2. Where is he?         (Location)

3. When he was?       (Time)

4. What is he doing?  (Activity)

Table1 illustrates some of examples for context and non context set.

Table1: Context and non context set of attributes

Set Of Context Attributes	Who	Where	What	When	Context	Example
Name ,Role	1	0	0	0	No	Rohan is Father
Name, Role, Time	1	0	0	1	No	Rohan is Father during Morning Hours
Name, Time ,Activity	1	0	1	1	No	Rohan is Father and Doing Yoga during Morning Hours
Name ,Time, Activity, Location	1	1	1	1	Yes	Rohan is Father and Doing Yoga during Morning Hours in the Living Room of his House.

                 Note : 1- Provides related information  0 : Does not provides the related information

The context attributes can be classified into two categories:

1.    Indispensable Attributes

2.    Dispensable Attributes

The indispensable attributes are those which provide all the mandatory information to describe context. All other attributes are dispensable attributes. But dispensable attributes adds more flavor to the indispensable attributes and will be useful for providing socialize and personalize services. Table2 illustrates some example of dispensable and indispensable attributes.

Table2: Dispensable and Indispensable Attributes

Context	Indispensable Attributes	Dispensable Attributes
Rohan is Father  and doing yoga during Morning Hours in the living room of his house	Name ,Role ,Activity ,Location	Nil
Rohan is Father , his health is normal and doing yoga during Morning Hours	Name ,Role ,Activity, Location	Health.
Rohan is Father , his sex is male ,health is normal ,physical posture is running and doing yoga during Morning Hours .	Name ,Role ,Activity, Location	Health ,Sex
Rohan is Father , his sex is male , health is normal , mood is normal, and eye sight is abnormal, and doing yoga during morning hours in the living room of his house.	Name ,Role ,Activity ,Location	Health, Sex, Eye Sight, Mood.

1.2 Properties of Contextual Information: The information related to context of an entity is termed as contextual information with respect to that entity. Some of the properties to list are as follows:

1.    Context information may be static or dynamic: The context of any entity may or may not change with respect to time. Context which represent the dynamic information are said to be dynamic context. Example: Age of person, User time point, users location, user friends, user   intention, user mood, etc.

The context which represent the static information are said to be static context.    Example: Name, Date of Birth, User Role, User Priority etc. However the static information like name, date of birth, priority of user can be represented as information with change frequency zero.

2.  Context information creates History: Since context varies from time to time a history of user/entity context will be created with respect to time. The history of context can be utilized for predicting the future context.

3. Quantity of Context information is large: It depends on uncertainty in measurements, applications of heuristics, assumptions for derivation and interpretation.

4. Incorrectness in Context Information: Due to inexact sensor information, measurement failures or due to wrong assumptions for derivation and interpretation.

5.    Multiple sources: Same information can be gathered in different ways. Eg: Location of a person (GPS,Position of device, WLAN)

6.     Relevance of Context Information : It depends on the following  two factors

1. Capturing time: The relevance will be maximum at capturing time .It decreases constantly. Example :Location of a mobile user and user interest on particular TV program.

2. Location: The temperature measured depends on the location .The relevance will be maximum at capturing place and decreases with distance.

7. Context Information is Multidimensional / Heterogeneous: The context information may be personal, physical, technical or social in nature.

8.    Context information is Distributed: The contextual information occurs everywhere / all the time. It is ubiquitous, pervasive and omnipresent.

9.    Context information may be imperfect: This is due to the incomplete data or inconsistency in data provided by the sensors.

10.  Context information is Unforeseeable: Since any information can be relevant as context the context information is unforeseeable.

1.3 Categories of Context: The context can be categorized into different types based on the information they carry. Following are some of the list of information based context.  

1. Computing Context: It provides the information related to the computing like processing speed, memory, free disk space, necessary software, OS services, internet connection, utilities, etc.

2.    Network Context: It provides the information related to networking like connection, bandwidth, LAN, WAN, WIFI, Bluetooth, signal strength, etc.

3.    User Context: It provides the information related to user like name, role, location, priority, activity, mood and other user’s information.

4.  Physical Context: It provides the information related to environment like light intensity, temperature, weather conditions, sound level, etc.

5.   Time Context: It provides the information related to time such as time of day ,week and Month.

6.  Sensor Context: It provides the information related to the different sensors like location, time, day, temperature, noise, light intensity, etc to perceive the context of environmental input. The sensor context includes the information like active, inactive, damaged, out of order, switched on/off, resolution, relevance, etc. The sensors provide the raw contextual data to the drivers or widgets attached to the sensors.

7.    Device Context: It provides the information related to the different actuators (TV, AC, FAN, Light, Mobile, etc) like device input and output capabilities, memory, software support, available services, service preferences etc., to provide meaning full service to the users. Devices are the service providers for users.

The context can be further divided into primary context and secondary context. The primary or low level context refers to the environment characteristics which can be gained directly from sensors, eg: location, time, nearby objects, network bandwidth, orientation, light level, sound and temperature. The sensors can either measure physical parameters in the environment or logical information gathered from the host (eg.Current time , GSM cell ,selected action) and the sensors are called physical or logical correspondingly. The secondary or high level context refers to a more abstract context, which is derived from the primary context: the user’s social situation, activity or mental state.

                                                                                                           ....Dr.Thyagaraju G S

References 

[1]   Albert Schmidt ,”Ubiquitous Computing – Computing in Context”,Ph.D Thesis Submitted to Lancaster University ,November 2002.

[2] DEY, A. K., AND ABOWD, G. D. 2000a. Towards a Better Understanding of Context and Context-Awareness. In Proceedings of the Workshop on The What, Who, Where, When, and How of Context-Awareness within CHI’00, 1-12.

[3]  Pascoe J : Adding generic contextual capabilities to wearable computers. In : Proceedings  of the Second International Symposium on Wearable Computers.(1998) 92 -99.