Mining Human Movement Patterns For Big Data For Healthcare Applications in Smart Home

International Journal of Trend in Scientific
Research and Development (IJTSRD)

International Open Access Journal
ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume - 2 | Issue – 3
Mining Human Movement Patterns ffor

Big Data for Healthcare Applications in Smart Home
Ishwaramma
Ishwaramma, Prof. Mohamed Rafi
Department off Studies in Computer Science & Engineering, University BDT College of Engineering
(A Constuent College of VTU, Belagavi), Davangere, Karnataka,, India
ABSTRACT
Nowadays, there is an ever-increasing

increasing migration of environment [2] [3]. In such transformation, millions
people to urban areas. Health care services are one of of homes are being equipped with smart devices (e.g.
the most challenging aspects that is greatly affected smart meters, sensors etc.) which generate massive
by the vast influx of people to city centers. volumes of fine-grained
grained and indexical data that can be
Consequently, cities around the world are investing analyzed to support health care services.
heavily in digital transformation in an effort to Advancement of big data mining technologies, which
provide healthier ecosystem for people. In this paper, provide means of processing huge amount of data for
we propose a model that utilizes smart home big data actionable insights, can aid us in understanding how
as a means of learning and discovering human activity people go about their life.
patterns for health care applications. We propose tthe
use of frequent pattern mining, cluster analysis and For example, monitoring the changes of appliance
prediction to measure and analyze energy usage usage inside a smart home can be used to indirectly
changes sparked by occupants’ behavior. determine the person’s wellbeing based on historical
data. Since people’s habits are mostly identified by
Keywords: Big data, smart cities, smart homes, smart everyday routines, discovering these routines allows
cities, health care applications, Behavioral Analytics, us to recognize anomalous activities that may indicate
Frequent
quent Pattern, Cluster Analysis, Incremental people’s difficulties in taking care for themselves,
Data-Mining,
Mining, Association Rules, Prediction such as not preparing food or not using shower/bath
[4] [5]. The underlying
ng correlation between appliance
INTRODUCTION usage inside the smart home and routine activities can
be used by health care applications to detect potential
Studies show that by year 2050, 66% of the world health problems. This is not only going to alleviate the
population will be living in urban areas [1]. The burden on health care systems, but also providing 24
demand for health care resources will be greatly hour monitoring service that automatically identifies
affected
ed by this vast influx of people to city centers. normal and abnormal behaviors for independently
This unprecedented demographic change places living patients or those with self-limiting
self conditions
enormous burden on cities to rethink the traditional (e.g. elderly and patients with cognitive impairments).
approaches of providing health services to residents. In existing system .studies show that by year 2050,
In responding to the new needs and challenges, cities 66% of the world population will be living in urban
are currently embracing massive digital areas .There is no chance to know about their health
transformation in an effort to support sustainable issues which may lead a problem the demand for
urban communities, and provide healthier health care resources will be greatly affected by this
vast influx of people to city centers. This Thi
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 3 | Mar-Apr

Apr 2018 Page: 2170
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
unprecedented demographic change places enormous particularly useful for tracking individuals suffering
burden on cities to rethink the traditional approaches from self-limiting conditions such as Alzheimer's,
of providing health services to residents. Parkinson's disease and clinical depression. This
paper explores this idea further and presents a new
LITERATURE SURVEY approach for unobtrusively monitoring people in their
homes to support independent living. The posited
Abdulsalam Yassine proposes a Health care services
system uses data classification techniques to detect
is one of the most challenging aspects that is greatly
anomalies in behaviour through personal energy usage
affected by the vast influx of people to city centers.
patterns in the home. Our results show that it was
Consequently, cities around the world are investing
possible to obtain an overall accuracy of 99.17% with
heavily in digital transformation in an effort to
0.989 for sensitivity, 0.995 for specificity and an
provide healthier ecosystem for people. In such
overall error of 0.008 when using the VPC Neural
transformation, millions of homes are being equipped
Network classifier.
with smart devices (e.g. smart meters, sensors etc.)
which generate massive volumes of fine-grained and
M. Shamim Hossain proposes the Smart, interactive
indexical data that can be analyzed to support smart
healthcare is necessary in the modern age. Several
city services. In this paper, we propose a model that
issues, such as accurate diagnosis, low-cost modeling,
utilizes smart home big data as a means of learning
low-complexity design, seamless transmission, and
and discovering human activity patterns for health
sufficient storage, should be addressed while
care applications. We propose the use of frequent
developing a complete healthcare framework. In this
pattern mining, cluster analysis and prediction to
paper, we propose a patient state recognition system
measure and analyze energy usage changes sparked
for the healthcare framework. We design the system
by occupants’ behavior Since people’s habits are
in such a way that it provides good recognition
mostly identified by everyday routines, discovering
accuracy, provides low-cost modeling, and is scalable.
these routines allows us to recognize anomalous
The system takes two main types of input, video and
activities that may indicate people’s difficulties in
audio, which are captured in a multisensory
taking care for themselves, such as not preparing food
environment. Speech and video input are processed
or not using shower/bath. Our work addresses the
separately during feature extraction and modeling;
need to analyze temporal energy consumption patterns
these two input modalities are merged at score level,
at the appliance level, which is directly related to
where the scores are obtained from the models of
human activities. For the evaluation of the proposed
different patients’ states.
mechanism, this research uses the UK Domestic
Appliance Level Electricity dataset (UK-Dale) - time
For the experiments, 100 people were recruited to
series data of power consumption collected from 2012
mimic a patient’s states of normal, pain, and tensed.
to 2015 with time resolution of six seconds for five
The experimental results show that the proposed
houses with 109 appliances from Southern England.
system can achieve an average 98.2 % recognition
The data from smart meters are recursively mined in
accuracy.
the quantum/data slice of 24 hours, and the results are
maintained across successive mining exercises. The
Michael Mackay proposes smart meters have allowed
results of identifying human activity patterns from
us to monitor consumers' energy usage with a high
appliance usage are presented in details in this paper
degree of granularity. Detailed electricity usage
along with accuracy of short and long term
patterns and trends can be identified to help
predictions.
understand daily consumer habits and routines. The
challenge is to exploit these usage patterns and
C.Chalmers, W.Hurst, was Proposes the smart meters
recognise when sudden changes in behaviour occur.
has allowed us to monitor consumers' energy usage
This would allow detailed, around the clock,
with a high degree of granularity. Detailed electricity
monitoring of a person's wellbeing and would be
usage patterns and trends can be identified to help
particularly useful for tracking individuals suffering
understand daily consumer habits and routines. The
from self-limiting conditions such as Alzheimer's,
challenge is to exploit these usage patterns and
Parkinson's disease and clinical depression. This
recognise when sudden changes in behaviour occur.
paper explores this idea further and presents a new
This would allow detailed, around the clock,
approach for unobtrusively monitoring people in their
monitoring of a person's wellbeing and would be
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 2171
homes to support independent living. The posited Fig. 1. Model: Mining Frequent Patterns and Activity
system uses data classification techniques to detect Predictions for Health care Applications in Smart
anomalies in behaviour through personal energy usage Homes
patterns in the home. Our results show that it was
possible to obtain an overall accuracy of 99.17% with Figure (1) represents the proposed model. It starts by
0.989 for sensitivity, 0.995 for specificity and an cleaning and preparing the data and then applying
overall error of 0.008 when using the VPC Neural frequent pattern mining for discovering appliance-to-
Network classifier. appliance associations, i.e., determining which
appliances are operating together. Then, it uses cluster
PROPOSED SYSTEM analysis to determine applianceto- time associations.
With these two processes, the system is able to extract
In this paper, we propose a model that utilizes smart the pattern of appliance usage which is then used as
home big data as a means of learning and discovering input to the Bayesian network for short-term and
human activity patterns for health care applications. long-term activities prediction. The output of the
We propose the use of frequent pattern mining, cluster system is utilized by specific health care applications
analysis and prediction to measure and analyze energy depending on the intended use. For example, a health
usage changes sparked by occupants’ behavior. Since care provider might only interested in knowing
people’s habits are mostly identified by everyday activities related to cognitive impairment where
routines, discovering these routines allows us to tracking the sequence of daily activities is crucial for
recognize anomalous activities that may indicate reminding the patient when abnormal behavior is
people’s difficulties in taking care for themselves, detected. Next subsection explains such processes and
such as not preparing food or not using shower/bath. briefly outlines the theoretical background. We
Our work addresses the need to analyze temporal propose a human activity pattern mining model based
energy consumption patterns at the appliance level, on appliance usage variations in smart homes. The
which is directly related to human activities. For the model which utilizes FP-growth for pattern
evaluation of the proposed mechanism, this research recognition and k-means clustering algorithms is
uses the UK Domestic Appliance Level Electricity capable of identifying appliance-to-appliance and
dataset (UK-Dale) - time series data of power appliance-to-time associations through incremental
consumption collected from 2012 to 2015 with time mining of energy consumption data. This is not only
resolution of six seconds for five houses with 109 important to determine activity routines, but also,
appliances from Southern England. The data from when utilized by health care application, is capable of
smart meters are recursively mined in the detecting sudden changes of human activities that
quantum/data slice of 24 hours, and the results are require attention by a health provider. We apply a
maintained across successive mining exercises. The Bayesian network for activity prediction based on
results of identifying human activity patterns from individual and multiple appliance usage. This is
appliance usage are presented in details in this paper significant for health applications that incorporate
along with accuracy of short and long term reminders for patients to perform certain activities
predictions. based on historical data. For added accuracy of the
system, the prediction model integrates probabilities
ARCHITECTURE DIAGRAM of appliance-to-appliance and appliance-to time
associations, thus recognizing activities that occur in
certain patterns more accurately.
ADVANTAGES
• Protect people’s privacy from being shared .
• To understand and predict their activities that could
indicate health issues.
CONCLUSION
In this paper, we presented a model for recognizing

human activities patterns from low resolution smart
meters data. Occupants’ habits and behavior follow a (NBiS), 2015 18th International Conference on, 9
pattern that could be used in health applications to 2015, pp. 273–279.
track the wellbeing of individuals living alone or
8. K. Jack and K. William, “The UK-DALE dataset,
those with self-limiting conditions. Most of these
domestic appliance-level electricity demand and
activities can be learned from appliance-to appliance
whole-house demand from five UK homes,”
and appliance-to-time associations. We presented
Scientific Data, vol. 2, no. 150007, 2015.
incremental frequent mining and prediction model
based on Bayesian network. In our current work, 9. J. Clement, J. Ploennigs, and K. Kabitzsch,
through experiments, we found that 24-hour period Detecting Activities of Daily Living with Smart
was optimal for data mining, but we built the model to Meters. Springer, Germany, 11 2014, ch. Advance
operate on any quantum of time. From the experiment Technology and Societal Change, pp. 143–
results we have demonstrated the applicability of the 160.[Online]. Available:
proposed model to correctly detect multiple appliance https://link.springer.com/chapter/10.1007/ 978-3-
usage and make short and long term prediction at high 642-37988-8 10
accuracy. 10. Q. Ni, A. B. Garca Hernando, and I. P. de la Cruz,
“The elderlys independent living in smart homes:
REFERENCES A characterization of activities and sensing
infrastructure survey to facilitate services
1. N. United, “World urbanization prospect.” United development,” Sensors, vol. 15, no. 5, pp. 11 312–
Nation, 2014. [Online]. Available: 11 362, 2015. [Online]. Available:
http://dl.acm.org/citation.cfm?id=308574. 308676 http://www.mdpi.com/1424-8220/15/ 5/11312
2. M. S. Hossain, “Cloud-supported cyber-physical
localization framework for patients monitoring,”
IEEE Systems Journal, vol. 11, no. 1, pp. 118–
127, March 2017.
3. M. S. Hossain, G. Muhammad, W. Abdul, B.
Song, and B. Gupta, “Cloud-assisted secure video
transmission and sharing framework for smart
cities,” Elsevier, Future Generation Computer
Systems Journal, April 2017.
4. J. Liao, L. Stankovic, and V. Stankovic,
“Detecting household activity patterns from smart
meter data,” in Intelligent Environments(IE), 2014
International Conference on, 6 2014, pp. 71–78.
5. A. Yassine, A. A. N. Shirehjini, and S.
Shirmohammadi, “Smart meters big data: Game
theoretic model for fair data sharing in
deregulated smart grids,” IEEE Access, vol. 3, pp.
2743–2754, 2015.
6. A. Yassine and S. Shirmohammadi, “Measuring
users’ privacy payoff using intelligent agents,” in
2009 IEEE International Conference on
Computational Intelligence for Measurement
Systems and Applications, May 2009, pp. 169–
174.
7. Y. C. Chen, H. C. Hung, B. Y. Chiang, S. Y.
Peng, and P. J. Chen, “Incrementally mining
usage correlations among appliances in smart
homes,” in Network-Based Information Systems

Mining Human Movement Patterns For Big Data For Healthcare Applications in Smart Home

Uploaded by

Copyright:

Available Formats

Mining Human Movement Patterns For Big Data For Healthcare Applications in Smart Home

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mining Human Movement Patterns For Big Data For Healthcare Applications in Smart Home

Uploaded by

Copyright:

Available Formats

International Journal of Trend in Scientific

Research and Development (IJTSRD)

Mining Human Movement Patterns ffor

Nowadays, there is an ever-increasing

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 3 | Mar-Apr

In this paper, we presented a model for recognizing

You might also like