Proceedings of the Second International conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2018)
IEEE Xplore Part Number:CFP18OZV-ART; ISBN:978-1-5386-1442-6
Context Aware Computing Systems: A survey
1
Ms. Pooja S. Gandodhar, Research Scholar, SKNCOE, SPPU, Pune
2
Dr. S.M. Chaware , PhD Guide, SPPU, Pune
 person, place or object that is considered relevant to the
Abstract— The term context has been studies in different
areas of Computer Science. Context-aware systems have
high demand in areas like Intelligent Environments,
Pervasive & Ubiquitous Computing and Ambient
Intelligence. Such systems gather data and adapting
system behavior accordingly using context information
like physical context, computational context, and user
context/tasks. Developing such context-aware applications
is inherently complex and hence should be supported by
adequate context information modeling and reasoning
techniques. In this paper the concept of context awareness
is discussed along with requirements for context
identification and formulation, procedure for converting
context modeling to intelligent actions and context
recognition techniques & algorithms are discussed. The
applications of context aware are reviewed.
Index Terms—Context, Context Awareness, Context modeling,
Context Aware Computing, Pervasive Computing
I. INTRODUCTION
One of the areas of pervasive (ubiquitous) computing is
context-aware computing. Mobility of devices and use of
services make context-aware systems a popular research field.
In real physical context means to use the data gathered from
sensors to the context-aware computing platform, which
includes: what to sense, how to acquire the information and
reasoning to that information to infer the context of a user. It is
highly needed that programs and services should respond
according to the user’s situation and behave the way she wants
these to be, i.e. services and systems should be more dynamic.
To identify the context the information which we require can
be captured in a number of ways, for example from user
profile information, network (to sense location, time, nearby
objects etc.), sensors (for activity) and other sources.
Many researchers have defined context as per their
understanding in an effort to consider a more general concept
of context. Schilit and Theimer [6] used the term context-
aware in 1994 and described as identities, location, objects and
nearby people [6]. In 1996 Brown defined context as the
elements that surround a user which a computer can identify
[5]. An often cited and quite generic definition of context is
that by Dey and Abowd: “Context is any information that can
be used to characterize the situation of an entity. An entity is a
978-1-5386-1442-6/18/$31.00 ©2018 IEEE
interaction between a user and an application, including the
user and applications themselves.” The context is useful and
people have worked on it, focusing on location mostly,
although the situation of an entity might consist time, location,
activity and the surrounding factors that might affect the
activity of an entity[6].
The term context has been classified into two categories
(physical and logical). Physical context can be determined by
hardware sensors and logical context is either given by the
user’s input or by capturing her interactions with the services
available [1][2]. For example through observing or analyzing
the user’s profile, activities, working routines, typing activity
etc. Most research in this area makes use of physical sensors
for light, movement, sound, temperature, touch, and of course
location. The logical sensors however provide related
information by reading user’s information from opened web
pages and other documents and also analyses user’s data
(interactions) and based on those interactions targets
advertising.
II. LITERATURE REVIEW
Following section includes discussion on concepts of context,
context awareness, and context aware systems.
Context The term Context is been studied by many
researchers. The most popular definition is information that
can be used to relate or characterize an entity [2]. Using
various definitions and interpretations “Context is any
information that can be used to characterize the situation of an
entity. A person, place, or object can become entity which is
considered applicable to the interaction between a user and an
application, including the user and applications
themselves.”[3][6]
Context Awareness was introduced Schilit and Theimer as
sentient [7]. “A system is context-aware if it uses context to
provide relevant information and/or services to the user, where
relevancy depends on the user’s task.[4] [9]”
Context-aware systems are able to adapt their activities based
on current context. This also increases effectiveness by taking
environmental context into account. Context aware systems
watch the environment all the time and propose suitable
suggestions to users so they can take necessary actions. For
example publishing a user's location to appropriate members
of a social network, and allowing retailers to publish special
offers to potential customers who are near to the retailers.
605
 Proceedings of the Second International conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2018)
IEEE Xplore Part Number:CFP18OZV-ART; ISBN:978-1-5386-1442-6

Context aware systems used variety of contexts like location,
environment, and device. A user’s current location to a device
was considered as context [10]. Ambient environmental
context includes the profile of an environment [11]. Systems
modeling device context includes adjusting screen brightness
based on the available battery power [12]. User context
includes her physical activity, her mode of transport and user
activity logs, usage styles for commute and shopping
references etc.
In context aware systems, context can be, raw data and also
meaningful information used to draw decisions. Data can
come from various sources. For defining context data need to
be preprocessed. When the data is diverse, it has to be
normalized and aggregated[13].
For assisting users in decision making from the context
inferences gathered data analytics techniques are of great help
[8].
The features extracted here are sent to machine learning
algorithms-supervised and unsupervised to generate training
data sets and construct a classification model.
Fig. 1. The data collected from input sensors are preprocessed
and forwarded to knowledge acquisition module. Next the data
generated in previous steps are submitted to rule based
classifier. Then data mining algorithms are applied over
training datasets to generate classifiers. These classifiers will
make context aware decisions
The following set of steps enlists procedure for converting
contextual information into intelligent actions.
1. Create computational models for context acquisition with
respect to tasks
2. Generalization of contextual characteristics based on the
tasks
3. Aggregation of gathered information to generate an
abstraction of context
4. Make choice of the algorithm or a set of algorithms to use.
5. Perform context recognition and derive inferences
6. Derive conclusions and assist the user with most suitable
decision acoordingly.

III. CONTEXT AWARE SYSTEM LAYERED IV. TECHNIQUES & ALGORITHMS

ARCHITECTURE The Context Modeling Schemes are discussed here.
A proposed basic multilayered architecture of context
information processing is shown in figure 1.
Sr. Type of Scope Limitations
No scheme
Input data .
{Sensor data,
activity logs, 1 Key- Simplest Different
historical data} Value Model context context types.
Modelin information to text Dependencies
Pre-processing g files & binary files. timeliness.
{Data
2 Mark-up Data modeling by Not able to
summarization,
Scheme using tags-XML reason,
feature selection,
Modelin Used for profile interoperabili
normalization}
g modeling ty and
Knowledge reusability
Acquisition
{Context 3 Graphic Model context with Interoperabili
correlations, al reasoning for large ty is difficult
association rules, Modelin datasets of
thresholds} g databases
Post Processing
{Rule based
classifications,
converted data } 4 Object Model context Suitable for
Based using hierarchical internal
Modelin class interfaces context
Data Mining
{Supervised/Unsuperv g modeling
ised learning, based on
performance code
evaluation} 5 Logic Context is Lack of
Context Aware
Based modeled using standardizatio
Decisions
Modelin rules suitable for n leads to less
{Classifiers, decisions,
g reasoning reusability
notifications}
Fig.1 A basic multilayer architecture of a context processing
system
The different modules of the proposed system are shown in

978-1-5386-1442-6/18/$31.00 ©2018 IEEE 606

 Proceedings of the Second International conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2018)
IEEE Xplore Part Number:CFP18OZV-ART; ISBN:978-1-5386-1442-6

6 Ontolog Context is Large of rules

y Based formulated from datasets
Modelin common entity increases the
g knowledge computationa
Ontologies are lly and time 5 Ontology It is based on It is not good
mostly chosen for consumption Based description logic, in finding
modeling context Reasoning which is a family missing
Method of logic-based values or
knowledge ambiguous
representations information
Table 1. Comparison of Context Modeling Schemes comparing

Table 1 listed various context modeling schemes out of which 6 Probabil Collect data from Need to
ontology based modeling seems to be well suited for defining istic different sources. explore with
context. However it as data size increases the retrieval of Logic It is commonly combining
context from this would be time consuming [14] [16]. used for activity multiple
recognition of contexts
context-aware
Sr. Type of Scope Limitation applications such
No scheme s as situation
1 Supervis Training examples Training recognition in a
ed are collected and process will smart home
Learnin labelled according define Table 2. Comparison of Context Reasoning Schemes
g to the expected efficiency
targets or results. Table 2 listed various context reasoning schemes for assisting
Generalization users based on context inferences [16]. The most appropriate
function scheme is supervised learning as training data sets are labeled
Widely used in according to expected targets. Also probabilistic logic scheme
mobile phone is suitable for context activity recognition and resolving
sensing and conflicts between different sources of data.
activity
recognition V. CONTEXT AWARE APPLICATIONS
The context aware applications can be found in many different
2 Unsuper It is not provided Learning application domains as below which are gaining intensive
vised with the expected time attentions worldwide.
Learnin results during the increases
g training process. with 1. Location based services
Due to no training hierarchical Physical services (multiplexes, gardens ...)
data, there is no learning Electronic services (scanners, routers ...)
error or reward 2. Using location to improve (network) services
signal to evaluate a Rerouting of incoming or outgoing communications
potential solution. 3. Using location to provide information
City tours
3 Rule Rules are usually generalizatio Nearby business advertisements
based structured in an IF- n and 4. Making others aware of user location
scheme THEN-ELSE numbers of Presence (individual)
format for rules Popularity, movement (group)
generating of high- 5. Security
level context User access grants based on location
information from 6. Location Recommendations in Location based social
low-level context. networks
Recommend locations using a user’s location histories
4 Fuzzy Usage of language Complicated and community opinions
Logic variables and the systems may Location bridges gap between physical world & social
ability to deal with require networks
vague systems several 7. Existing Solutions
iterations to Based on item/user collaborative filtering
find a set Similar users give the similar ratings to similar items

978-1-5386-1442-6/18/$31.00 ©2018 IEEE 607

 Proceedings of the Second International conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC 2018)
IEEE Xplore Part Number:CFP18OZV-ART; ISBN:978-1-5386-1442-6

[14] Xin Li, Martina Eckert, José-Fernán Martinez and

VI. CONCLUSION
This survey has analyzed the concept of context and context
awareness. An analysis of context modeling and reasoning
techniques is done. This paper also enlisted popular context
aware systems applications.
The future trend of context-aware applications will be based
on social context. Social context should not only focus on the
interactions among people but also among machines and even
between the people and the machine (IoT & IoE). The
significant outcome of context aware application is not only
acquisition of data but also derive inferences from those data.
Gregorio Rubio, “Context Aware Middleware
Architectures:Survey and Challenges” Sensors 2015
[15] C. Perera, A. Zaslavsky, P. Christen and D.
Georgakopoulos, "Context Aware Computing for The Internet
of Things: A Survey," in IEEE Communications Surveys &
Tutorials, vol. 16, no. 1, pp. 414-454, First Quarter 2014.
[16] Punnarumol Temdee, Ramjee Prasad, Book: “Context
Aware Communication & Computing: Applications for Smart
Enviornment”, Springer Series in Wireless Technology 2018

VII. REFERENCES
[1] Unai Alegre, Juan Carlos Augusto, Tony Clark,
Engineering Context- Aware Systems and Applications: A
survey, The Journal of Systems & Software (2016),
doi:10.1016/j.jss. 2016.02.010
[2] K.P. Subbu, A.V. Vasilakos, Big Data for Context Aware
Computing – Perspectives and Challenges, Big
Data Res. (2017), https://doi.org/10.1016/j.bdr.2017.10.002
[3] Eunhoe Kim and Jaeyoung Choi, “A Context Management
System for Supporting Context-Aware Applications”, 2008
IEEE/IFIP International Conference on Embedded and
Ubiquitous Computing
[4] Simon Elias Bibri & John Krogstie, “Big Data and
Context–Aware Computing Applications for Smart
Sustainable Cities” 2016 - researchgate.net
[5] A.K. Dey and G.D. Abowd: Towards a Better
Understanding of Context and Context-Awareness. In
Proceedings of CHI 2000 Workshop on the What, Who,
Where, When and How of Context Awareness, 2000
[6] B. Schilit, N. Adams, R. Want, Context-aware computing
applications, in: Mobile Computing Systems and Applications,
1994. WMCSA 1994. First Workshop on, IEEE, 1994, pp. 85–
90.
[7] M. Chen, S. Mao, Y. Liu, Big data: A survey, Mobile
Networks and Applications 19 (2) (2014) 171–209.
[8] H. Hu, Y. Wen, T.-S. Chua, X. Li, Toward scalable
systems for big data analytics: A technology tutorial, IEEE
Access 2 (2014) 652–687
[9] R. Want, A. Hopper, V. Falc˜ao, J. Gibbons, The active
badge location system, ACM Trans. Inf. Syst. 10 (1) (1992)
91–102.
[10] R. Want, B. N. Schilit, N. I. Adams, R. Gold, K. Petersen,
D. Goldberg, J. R. Ellis, M. Weiser, An overview of the
parctab ubiquitous computing experiment, IEEE Personal
Communications 2 (6) (1995) 28–43.
[11] D. Siewiorek, A. Smailagic, J. Furukawa, A. Krause, N.
Moraveji, K. Reiger, J. Shaffer, F. L. Wong, Sensay: A
contextaware mobile phone, in: International Symposium on
Wearable Computers, 2003.
[12] Bibri SE. The human face of ambient intelligence,
cognitive, emotional, affective, behavioral, and conversational
aspects. Berlin: Springer-Verlag; 2015.
[13] Bibri and Krogstie, “The core enabling technologies of
big data analytics and context‑aware computing for smart
sustainable cities: a review and synthesis”J Big Data (2017)

978-1-5386-1442-6/18/$31.00 ©2018 IEEE 608