147
slope, step, and surface, and dynamic accessibility which could be man-made, such
as construction. Geometry of points composes coordinates of decision points and
POIs.
Table 8.2 shows the requirements of model-centric navigation from a data per-
spective (point) for the purpose of computation. This table summarizes the navi-
gation requirements of people with special needs. When navigation is in outdoor,
point geometry composes coordinates of intersections and POIs. Topology of points
contains connectivity of intersections through road/sidewalk segments. Attributes
include common propertie, such as name and type (e.g., restaurant, bookstore),
static accessibility, such as accessible pathway and accessible entrance, natural dy-
namic accessibility, such as accessible pathway condition due to snow/rain, and
man-made dynamic accessibility, such as construction.
When navigation is in indoor, point geometry composes coordinates of decision
points and POIs. Point topology contains connectivity of decision points through
hallway segments. Attributes include common properties, such as number and type
(e.g., room, restroom, drinking fountain), static accessibility, such as door, area,
protruding objects (e.g. fire extinguisher), room, width, grab bar, restroom, width
of sink, depth of sink, height of sink, height of drinking fountain, and man-made
dynamic accessibility, such as construction.
Table 8.3 shows the requirements of experience-centric navigation from a data
perspective. Recommendations for both outdoor and indoor navigation include POI
(name, address, accessibility) and route/direction (name, accessibility). Member’s
profile, which is used for both outdoor and indoor navigation, includes group cat-
egory (i.e., mobility impaired, visually impaired, cognitively impaired, elderly) and
preferences.
8.7 Model-Centric vs. Experience-Centric Navigation
Fig. 8.13
Issues affecting
trust of people on social
navigation networks
Trust:
High
Trust:
Low
Inaccurate/Non-relevant/
Non-personalized
Recommendations
Accurate/Relevant/
Personalized
Recommendations