Geography Reference
In-Depth Information
as professionally collected data, a relatively significant component of uncertainty
enters into the measurement chain. For this reason, it is necessary to firstly
disassemble the nature of voluntary noise data and compare it with the standard
noise data that are commonly processed today.
Volunteered Input Data Quality and Their Dependence
on User Behavior
The quality of data obtained during voluntary measurements is entirely dependent
on volunteer behavior during measurement. Generally, it is possible to distinguish
between two kinds of data obtained by this so called crowdsourcing:
• Data obtained through dedicated activity in the volunteer
s free time
(macrowork). In the case of environmental noise measurement, this means
getting near to the standards and methodologies for measuring noise as possible,
and therefore performing relatively long static measurements. Resulting data is
standard in nature and certified procedures can be used for processing and
interpretation. The only major difference from volunteered data compared
with that obtained by a certified device is a lower accuracy due to use of less
accurate sensor. Feick and Roche (
2013
) calls it quasi-scientific data, but the
question here is “what is scientific and what not?”.
• Data obtained during normal (informal) human activities (microwork). Using
this method it is possible to collect much more data than in the previously
described case, since it does not require any special volunteer time (if the mobile
collecting app is properly designed and requires only a minimum of user
operations). When a sufficient number of measurements are collected and the
user
'
s attention on the measurement process is sufficient, this method also
collects fairly reliable data on environmental conditions to which the volunteer
is exposed in his/her real life.
'
s
reader. The second type of data is much more prone to interpretation errors due to
its higher degree of its uncertainty and therefore graphical presentation of this data
requires a more complex post-processing. It is thus necessary to apply the theory of
uncertainty, which may greatly affect the resulting cartographic representation.
It is problematic to combine these types of data, or it may confuse the map
'
