Information Technology Reference
In-Depth Information
have an ID field and any data with the same ID coming from the same data
source are assumed to be connected.
The data types are used to determine which visualization plugins can visualize
what data. A clock plugin for timestamps will only be matched to data of that
type, and plugins that visualize geographical locations will only be matched to
data containing geographical information, etc. The system itself does not need
to understand the data types, the only requirement is that the data source and
the visualization plugins agree on the name and format of the data type. This
means that it is possible to add plugins that handle new data types that the
Dashboard
system has never seen without changing anything in the system.
The interface of the data source components is very simple. They are ex-
pected to have two slots:
PluginName
and
DataValues
.The
PluginName
slot
should contain a string with the name to display to the user when necessary.
The
DataValues
slot should contain the data in XML.
When doing a mash-up of several data sources, the
Dashboard
is used to
choose how to join the different data sets. Data fields in separate data sets that
have the same data types can be joined. An example: we might have data with
latitude, longitude, time-of-day and text in our
Twitter
data source; data with
latitude, longitude, time-of-day, and speed in our
Probe car
data set; and data
with time-of-day and temperature in our
Weather sensor
data set. Then the
Twitter
and
Probe car
data sets could be joined on the time-of-day and on the
location, and all three data sets could be joined on the time-of-day. Selecting a
geographic area using some visualization component would then remove subsets
of both the
Twitter
data and the
Probe car
data (data points outside the selected
area), but would not affect the
Weather sensor
data. Selecting only data from
the night time would affect all data sources, and selecting only data containing
the word “snow” would only affect the
Twitter
data.
2.3 Visualization Components and Their Interface
The
Digital Dashboard
hides the properties of the data and the existence of
and interaction with any other visualization components from each visualization
component. The parts of the visualization component that interact with the rest
of the
Dashboard
are thus simple, and more effort can be spent on the actual
visualization. Visualization components are expected to have the following slots:
PluginName
, should have a string indicating what the plugin should be called
in the user interface, e.g. when selecting which plugin to use for certain data.
RequiredDataFields
, a slot with XML describing what data fields this plugin
wants. A plugin showing dots on a map for locations of sent text messages and
displaying the message text when the mouse pointer is over the message location
could specify that it wants a latitude, a longitude, and a text field. Fields can
be optional, e.g. it could be OK that some items have no text but all items must
have latitude and longitude. A plugin can accept several different sets of data
fields, for instance a plugin showing road segments could accept fields with start
and end points or fields with a start point, a direction, and a segment length.