Mining Network Relationships in the Internet of Things

croutonsgruesomeΔίκτυα και Επικοινωνίες

16 Φεβ 2014 (πριν από 3 χρόνια και 3 μήνες)

67 εμφανίσεις

Mining Network Relationships in
the Internet of Things
PAT DOODY,
DIRECTOR OF THE CENTRE FOR INNOVATION IN DISTRIBUTED SYSTEMS
(CIDS)
INSTITUTE OF TECHNOLOGY TRALEE
ANDREW SHIELDS
IRC FUNDED RESEARCHER
CENTRE FOR INNOVATION IN DISTRIBUTED SYSTEMS
INSTITUTE OF TECHNOLOGY TRALEE
Overview
Reality Mining?
Reality mining Applied to IoT
Driving Applications
Conclusions

Reality Mining
“…The collection and analysis of machine-sensed
environmental data pertaining to human social
behaviour …” [1]
.. Extracting information from real world sensor
data
With the goal of identifying predictable patterns of behaviour.
It was declared to be one of the "10 technologies
most likely to change the way we live" by Technology
Review Magazine 2008.
[1] N. Eagle and A. Pentland, “Reality mining: sensing complex social systems,”
Ubiquit Computing, pp. 255-268, 2006. Reality Mining

Reality Mining applied urban transport networks
It is possible to infer an individual’s
Daily commute to work
Amount of time spent at work, at home and traveling
Allowing individuals to make better traveling
decisions.
Provides information which will be used to
proactively manage the transportation network.
Several clustering algorithms base on artificial
intelligence and statistical analysis will need to be
considered and evaluated
Adaptive Resonance Theory,
Citysense
Shows the overall activity
level of the city,
Highlights top activity
hotspots in real-time.
Then it links to Yelp and
Google to show what
venues are operating at
those locations.
https://www.sensenetworks.com
Reality mining Applied to IoT
Data mining as applied to “business intelligence”
applications may play a role
Techniques currently applied to understanding
human behaviour and interactions may be
applicable to IoT systems.
Reality Mining is one such technique.

Small-world networks
Objects may only use knowledge of their
own acquaintances, to collectively construct
paths to the target.
“six degrees of separation” found by the
social psychologist Stanley Milgram
Why should this type of decentralised
routing so effective?
Contains a “gradient” that guides messages toward the target.
Clustering
A common property of human social networks are
cliques, circles of friends or acquaintances
This inherent tendency to cluster is quantified by the
clustering coefficient [Watts and Strogatz (1998)].
Nodes that are clustered together can easily
communicate with each other.
Previous research in this area (Ghiasi, et al. 2002)
has studied the theoretical aspects of this problem
Applications to energy optimisation.

Degree Distribution
Nodes in a network typically do not all have the same
number of links, or degree.
For a large number of networks
The World Wide Web [Albert et al. (1999)],
The internet [Faloutsos et al. (1999)]
metabolic networks [Jeong et al. (2000)],
The work listed above assumes a static network
topology
Complex IoT networks will continuously changing
over time.

Algorithm Considerations
Algorithms must take into consideration the
characteristics of networks
Energy,
Computation constraints,
Network dynamics, and faults.
K-Nearest Neighbor Algorithm
ART1
Weighted Regression
Case-based reasoning

Enabling technology and infrastructure
The widespread adoption of the Internet of Things will take time
First: in order to connect everyday objects item identification is
crucial.
Radio-frequency identification (RFID) offers this functionality.
What if we can’t identify an object can we infer the object type from its behaviour?
Second: the ability to detect changes in the physical status of
things, using sensor technologies.
Embedded intelligence in the things themselves can further enhance the power of the
network
Third: advances in miniaturisation and nanotechnology mean
that smaller and smaller things will have the ability to interact
and connect.
A combination of all of these developments will create an
Internet of Things that connects the world’s objects in both a
sensory and an intelligent manner.

Mining the IoT Social Network
Relationships between smart object in an IoT
network
May have similar properties to humans interacting in a social
environment.
When smart objects participate in context-aware
applications
Changes in their real-world environment impact on underlying
networking structures.
Vast amounts of data being generated by smart
objects
Modelled and applied to complex IoT networks
Mining the IoT Social Network
Randomness (entropy)
Inherent in human social networks
Entropy of a smart object may be used as a metric


Mining the IoT Social Network
Dyadic Inference.
Human social networks respond to surrounding
social environment
Smart objects may exhibit similar dyadic properties.
From these properties it may be possible to infer
Relationships between multiple smart objects
based on patterns in proximity data.
Smart objects related in such a manner may responds to
environmental stimuli

Why do we care?
Social Science
Social Network Analysis
Behavioural Modelling
Human Mobility
Systems Research
Transportation
Environmental Modelling
Healthcare

Driving Applications
User-Generated Content is a core aspect of the Web
online social networks
Blogs
wikis,
Forums
One of the most successful services allowing this is
Twitter:
Possibility is the development of Things-Generated
Content where Things (instead of human beings) are
provided with "tweet-capabilities"

CIDS Research
Altobridge
Remote Villages
Reality Mining applied to mobile networks
Classifying user groups
Predicting network usage patterns
Using neural Network and other techniques
Problem: Data backhaul uses satellite

Challenges
Large Datasets
Wal-Mart: 100-400 GB/day of RFID data
CERN LHC: 40 TB/day
Storage is cheap!
Stream data mining
Data Mining Agents

Challenges
Abstraction
Low level details
Parallelism
Task distribution
Load balancing
Fault tolerance
Google (MapReduce ), Apache (Hadoop, Mahoot,
Zookeeper ) etc.
Frameworks to support distributed computing on large data
sets on clusters of computers

Challenges
Privacy
Right to possess data
Control the use of data
Right to distribute or dispose of data
Conclusions
The Internet of Things has great promise
Business, policy, and technical challenges must be
tackled before these systems are widely embraced.
Early adopters will need to prove that the new sensor
driven business models create superior value.
Industry groups and government regulators should
study rules on data privacy and data security,
particularly for uses that touch on sensitive consumer
information.
Software to aggregate and analyse data, must improve
to the point where huge volumes of data can be
absorbed by human decision makers or synthesised to
guide automated systems more appropriately.
Conclusions
On the technology side, the cost of sensors and
actuators must fall to levels that will spark widespread
use.
Networking technologies and the standards that
support them must evolve to the point where data can
flow freely among sensors, computers, and actuators.
Within companies, big changes in information patterns
will have implications for organisational structures, as
well as for the way decisions are made, operations are
managed, and processes are conceived.
Product development, for example, will need to reflect
far greater possibilities for capturing and analysing
information.