Abstract – Environmental pollution is a
major concern in developing countries. Ever
since industrialisation has begun, the percentage
of toxic gases in air has increased tenfolds. There
is a constant need for monitoring the increase in
toxic gas concentration and take certain steps to
combat the threat to environment. The existing
systems are restricted to hardware components
which only deals with toxic gas detection and
also do not have a mobile persepective. This
paper proposes an architecture that includes a
hardware device for detection of toxic gases as
well as a software system. This paper explains
the use of two data mining algorithms ,namely,
Bayes Theorem and K-Nearest Neighbor for
providing a safety recommendation solution
to the user. Experimental results show that the
proposed system is efficient and feasible in real
time environment.
Index Terms – IoT, sensor devices, , on-chip
sensors, bayesian process, k nearest neighbor,
human interface design.
I. INTRODUCTION
Internet of Things (IoT) is an emerging field
comprising of various sensory devices, which
are being applied to tackle environmental as
well as health issues. Environmental sensors
such as temperature, humidity etc. can be
deployed in outdoor and indoor locations.
This kind of network targets short distance
transmission of data. Main characteristics of
these sensors include low power consumption
and low cost. While designing such a sensory
system, following aspects need to be taken into
consideration:
- Sensor Potential
- Location (indoor/outdoor)
- Application (deployment)