Architecture for Including Personalization in a Mobile GIS via Semantic Web Techniques - Geographical Information Systems: Trends and Technologies - page 234

map data; 3) performs geographic operations; and 4) offers personalized
information automatically considering geographic, user and temporal
context.
First of all, we integrated the library of spatial functions described in
the previous section with the personalization algorithm in order to be able
to calculate distances and travel times. This algorithm uses the ontology
model, which is stored in the smartphone SDcard as a SQLite database. The
algorithm identifi es the most attractive POIs for a given user and situation
and sends them to the routing algorithm, which uses vector map data stored
previously in the smartphone in order to provide a suitable path to go from
one single POI to another.
In order to display geographic data, we used osmdroid library
(OpenStreetMap 2012), which provides tools to work with OpenStreetMap
data in an Android application. It allows displaying maps, getting the
coordinates when the user touches the screen and drawing lines and images
over the map. Hence, we implemented different functionalities to display
and interact with geographic data in the device.
Firstly, we developed a cache of maps from scratch in order to store
raster and vector map data downloaded from OSM server in order to avoid
memory limitations. The reason is that map tiles are displayed as raster
images with a high memory consumption. Since smartphone applications
have very limited memory by application, every time a zoom or a scroll is
done on the smartphone screen, the memory of the cache must be cleared.
The created cache of maps for Itiner@ application avoids such problem.
Secondly, we implemented an additional layer over the map view in
order to display the points of interest as clickable pushpins and to draw
the path between these points as blue lines.
Thirdly, we implemented a tool that allows the user to touch screen of
the smartphone to make the device aware of the coordinates of the map. By
doing so, the user can inform the application about his/her current location
when there is not GPS or 3G connection.
Note that the application connects to OSM Server to download the
required tiles and vector data before starting the route. This task should be
done when the user is at home or in a place with high bandwidth.
Proof of Concept
To show that the proposed theoretical approximation can be transferred to
a real application, we achieved to develop a proof of concept application
that could be the seed for a true intelligent travel guide able to work without
network connection, that allows people to use it anywhere and adapted to
every single user. Figure 10 shows some screenshots of the implemented
Android application.