Computational Aspects: How Landmarks Can Be Observed, Stored, and Analysed - Landmarks: GIScience for Intelligent Services - page 137

Tabl e 5. 1 Properties and measurement of visual, semantic, and structural attraction

Attractiveness

Measure Property Measurement Measure

p v f D R xjx 2 façade

s v is D .p v f C
p v sf Cp v sd C
p v c Cp vv /=5

Visual
Attraction

Façade

Shape

p v sf D height=width
p v sd D .area mbr˛/=area mbr

Color p v c D ŒR;G;B

p vv D P xjx visible Visibility

Semantic
Attraction

Cultural and
Historic

p sec D Œ0; 1 s sem D

.p sec Cp sem /=2

Explicit Mark p sem D Œ0; 1

Structural
Attraction

Nodes p stn D i Co s str D

.p stn Cp stb /=2

Boundaries p stb D cell size form factor

Adapted from [ 35 ]

Semantic attraction is made up of two measures:

￿ Cultural and historic importance: This property reflects whether an object is
culturally or historically important and, thus, becomes attractive. In the simplest
case, this is a Boolean value of true or false, but it may be refined, for example,
by measuring importance on a scale of 1 to 5.
￿ Explicit marks: An object may have explicit marks, such as signs on the front of
a building. These signs explicitly label an object communicating its semantics.
Again, this property is measured by a Boolean value.
Structural attraction is also measured using two properties:

￿ Nodes: here, nodes are the intersections in a travel network. The structural
importance of a node is defined by the number of incoming and outgoing
edges (the node degree). This could be further weighted by accounting for the
importance of these edges, for example, ranking highways higher than footpaths.
￿ Boundaries: boundaries separate two or more areas in geographic space. The
structural attractiveness of a boundary is linked to the effort that is required to
cross it. The higher its resistance, the higher its attractiveness. However, recent
findings show that this may be too simplistic an assumption [ 48 ] . The boundary
measure is the product of cell size and form factor in the formal model.

Tab le 5.1 summarizes the properties and measurement of the different kinds of
attractiveness used in Raubal and Winter's formal model.
The three individual attractiveness measures need to be combined in order to
calculate an overall attractiveness of a geographic object, which then gives the
landmark salience. Raubal and Winter's model uses a weighted sum to this end
(see Eq. 5.1 ) . The different weights can be adapted to reflect varying importance
of the individual attractiveness measures in different contexts, user preferences, or
applications. The original model sets all weights as equal, which is sensible if no
further information about each measure's influence is known.