two years, since 1988, a small but international group of academic and industry
researchers has been quietly exchanging their advances in spatio-temporal database
technology. Who are they? What edges are they exploring now? Is a graduate degree
required to comprehend their work? I had plenty of time to contemplate these questions
on the long series of flights between San Francisco, California, and Santorini
Island, Greece — site of the Eighth International Symposium on Spatial and Temporal
Databases in July. The trip was well worth the expense. Except for solitary
representatives from companies such as ESRI,
Oracle, and Logical Information
Machines, Inc., the attendees were professors and doctoral students driven more
by curiosity than the bottom line. If you emulate them, your curiosity may also be
piqued by questions such as:
This month, Net Results surveys several active spatio-temporal thinkers and
discusses their latest discoveries in this complex, highly technical domain. Rather
than attempt to cover any of their ideas in depth, we’ll just glance at the
highlights, like sampling a box of mixed chocolates, in hopes of conveying the
general flavor of today’s spatio-temporal research community.
- How can a database efficiently find the periods of overlap of time
intervals having vague or unknown beginnings or endings?
- Given a historic collection of moving points, how can the database identify
the area and instant of greatest point density?
- Of the many available indexing strategies, which will perform best
with very large or volatile spatiotemporal datasets?
- How can databases identify outliers when statistically analyzing point data?
- What kind of handheld PDA interface best supports the combination of LBS and
an event notification system for tourists?
- How can a central map server predict where a location-aware vehicle is going
in order to preemptively send the appropriate map updates?
LBS: Location-Based Services
PDA: Personal Digital Assistant
TIP: Tourism Information Provider