Database Reference
In-Depth Information
This data exploration has been done with a visualization tool. The user would
have also been able to perform the same extraction with a textual tool, such
as SQL queries. The only difference is that a textual tool would not have led
the user to the idea of exploring long flight duration in order to extract military
aircraft. Only with the incremental trajectory exploration can the user discover
the valid requests for this data set. In a sense, the user explores the data set, and
at the same time, explores the request to perform. Even if this process is efficient,
the direct manipulation cannot be automatic. Analysts need tools to enhance
their exploration capabilities. Therefore, extended work will be presented in the
following sections.
12.5 Event Extraction
There is a class of problems where analysts need to determine places in which
movement events (m-events) of a certain type repeatedly occur and then use
these places in further analysis. The relevant places can only be delineated by
processing movement data, that is, there is no predefined set of places (e.g.,
compartments of a territory division) from which the analyst can select places of
interest. The relevant places may have arbitrary shapes and sizes and irregular
spatial distribution. They may even overlap in space; therefore, approaches
based on dividing the territory into nonoverlapping areas, as in
Andrienko and
Andrienko
(
2011
), are not appropriate. In this section, we analyze one-day
record of aircraft trajectory with a visual analytics procedure for place-centered
analysis of mobility data (
Andrienko et al.
,
2011c
). The procedure consists of
four steps: (1) visually supported extraction of relevant m-events, (2) finding
and delineating significant places on the basis of interactive clustering of the
m-events according to different attributes, (3) spatio-temporal aggregation of the
m-events and movement data by the defined places or pairs of places and time
intervals; (4) analysis of the aggregated data for studying the spatio-temporal
patterns of event occurrences and/or connections between the places.
12.5.1 Analyzing Flight Dynamics in France
We shall apply our visual analytics procedure to ATC data with the following
goals: (1) Identify the airports in use. (2) Investigate the temporal dynamics of
the flights to and from the airports (i.e., landings and takeoffs). (3) Investigate
the connections among the airports, the intensity of the flights between them,
and their distribution over a day.
It may not be obvious to the reader why the airport areas need to be determined
from the data instead of using the official airport boundaries, which should be
known. The problem is the low temporal resolution of the data. For many flights,
the first recorded positions lie outside the boundaries of the origin airports and/or
