top of page
This site was designed with the
.com
website builder. Create your website today.Start Now

ABOUT

 

With the rapid improvement in smartphone technologies, we have been able to fit sensors, memory and wireless capabilities into small devices and see them be carried by vast majorities of the world. A data scientist like a GIS expert would be excited by this phenomenon because of the potential for us to mine large amounts of data like never before. 

 

Campbell, Andrew T., et al. "The rise of people-centric sensing." Internet Computing, IEEE 12.4 (2008): 12-21.

 

This prospect however brings a few considerations that we aim to study through our meta-analysis on the data collected by the participants of SGC 2015. We particularly look at trends in how the data was collected and how much of the collected data is truly useable for any form of geospatial analysis.

 

 

In this competition, 2 different sensors are used. Both collect the following data:

  • Temperature

  • Light

  • Noise

  • Carbon Monoxide Levels

  • Date & Time

All the collected data is stored online, to allow crowd sourcing of data for analysis.

 

GeoSense

 

It is an environmental sensor that is designed to collect data at a single location over long periods of time. Data points are added through ArcGIS Collector, a seperate mobile application. 

 

Advantages:

  1. Suitable for analysing data over time periods.

 

Disadvantages:

  1. Unsuitable for collecting data over large areas.

  2. Participants have to actively interact with sensor, and have to endure interruptions.

  3. Participants are aware of sensors, possible sampling biasness.

 

 

SENSg

 

SenSG is designed to collect data on the move; participants would walk around in with the sensor around their neck. As such, it is much smaller and lighter than GeoSense, and also has a built in GPS that will record the location of data point. 

 

Advantages:

  1. Suitable for collecting data over large areas.

  2. Does not require constant participant's interaction, reduce impact on participant's experience.

  3. Participants are less aware of sensor, less biased sampling.

 

Disadvantages:

  1. Unsuitable for analysing data over time periods.

  2. Privacy concerns.

 

 

Our work will be able to provide an evaluation to the data used by others working on this data set, analyse the methods used to collect this data and provide suggestions on how we can effectively involve the public in our data collection efforts. The data we have available from SGC 2015 is very apt as most of the participants are by no means experts at geospatial analysis and this entire data collection process behaves like a simulation for a project in which we involve the public in crowd sourcing our data.


Fortunately due to the structure of this competition, we have access to data from all the other groups. This saves us from a common problems faced by meta-analyses that have to do with availability of information such as publication bias and database bias.

bottom of page