- Supported by ARL, ORAU Ralph E. Powe Junior Faculty Enhancement Award, Texas ARP (PI, 003594-0010-2006), TxDOT (0-6432, 0-6767), NGA, NSF (co-PI, CI-TEAM0636421 ; PI, CNS-0709285 ; PI, IIS-0844342 , report ; PI IIS-1017926 )
Geo-sensor Network Databases:
The success of the ongoing sensor technology revolution depends critically on the flexibility and
ease with which people can fuse, query, and make sense out of the data provided collectively in
real time by networked sensors. We have several projects on large scale sensor deployment, education, and geo-sensor stream data management. Please visit our TEO website.
Departments of Transportation (DOT) and other government agencies are consistently trying to analyze the
travel behavior of individuals to help with urban planning. The placement of a Global Positioning System (GPS)
in a users' vehicle, and a written log classifying each trip can provide such information. One of the downfalls of a
written log is the high probability of errors. This inaccuracy can usually be attributed to the individual making incorrect
categorizations, or failing to make an entry. Written logs also tend to inhibit participation in studies due to the tenacity
needed by the participant. The goal of our project is to leverage algorithms and methodologies designed in data mining
community to tackle transportation problems.
Spatio-temporal Data Managment and Mining:
Advanced data collecting tools such as earth's observating system and global positioning system are accumulating
increasingly large spatial datasets. Since 1999 NASA's earth's observating system (EOS) has been studying changing
environment of the Earth and producing more than a terrabyte of data a day. Now more powerful, reliable, and
inexpensive location-enabled mobile devices are generating large geo-referenced datasets. These spatio-temporal datasets with
explosive growth rate requires new data models and query processing algoithms. They are also considered as nuggets of
valuable information and emphasize the need for an automated discovery of spatio-temporal knowledge.
Vertex Similarity from Graphs:
Privacy in Location-based Services
Evaculation planning and Vector Map Compression