Geoscience Reference
In-Depth Information
We are analyzing ways to provide positional privacy in the system, and
masking of user identities, preferring an opt-in approach to self-
identification, being aware of the many issues associated with motivations
of end-users and concerns about negative social dynamics. A major ongo-
ing effort is the integration of time stamps and temporal relevance into the
system.
3- Future Plans
At present, our two primary sources of user-contributed geospatial infor-
mation are the campus alert system and pedestrians who transit campus on
daily basis. We plan on growing the community of end-users and geospa-
tial data contributors by advertising the presence of our system through our
local campus disability services office and through the local campus plan-
ning office, which routinely provides information about sidewalk closures
associated with construction. We also plan on integrating other auditory
and haptic cues for obstacles and hazards, based on earlier work reported
by Rice, Jacobson, and Golledge (2005). We also intend to refine the tem-
poral aspects of the methodology described here. End-users typically lack
an understanding of the temporal dynamics of obstacles or hazards other
than a present-tense existence, i.e., “there is a hazard here right now”. In
many VGI-based systems, end-users don't have a way of specifying a
temporal end-point for their contributions and their appear to be few
resources directed at follow-up. A few authors, including Goodchild and
Glennon (2010) have discussed temporal issues, noting the primacy of
temporal relevance and the benefit of VGI, while maintaining a balance
between errors associated with false positives and false negatives. A gen-
eral approach to filtering and managing end-user contributions is to treat
the most recent update as the most relevant, and to phase out contributions
after a set period of time. This general approach, however, presents a num-
ber of problems when updates come from sources where material has been
rebroadcast or repackaged by news aggregation websites, causing it to
appear more recent and therefore more relevant. We are still working on
temporal issues and hope to have a better way of defining relevance and
end-points for contributions.
Other future goals for this research include the refinement of the method-
ology to deliver in-situ obstacle or hazard information to blind and visu-
ally-impaired individuals, and to improve the locational aspects of the
