Graduate Student Profiles
Kevin Ross
Kevin
Ross (kevin [at] ksross.com) graduated with his Master's Degree in
Geography in the summer of 2010 under the advisement of Dr. Alan M.
MacEachren. Kevin's thesis focused on geocollaboration, specifically
how teams of people query structured spatial data in a same-time,
different-place geocollaboration environment. For this research, he
developed a conceptual framework for how to support this type of collaboration,
programmed a web-based software application based on this framework,
then conducted a user study with this software to analyze how people
collaboratively query spatial data. Kevin also worked on other GeoVISTA
projects during his two years as a research assistant. He contributed
to the development of the original prototype of SensePlace, a tool
to support rapid web document acquisition and contextualization. The
main project Kevin worked on as a research assistant, and continues
to work on, is CrimeViz -- a web-based map application that supports
exploration and sensemaking of criminal activity in space and time.
More about Kevin's Thesis Research
SQSynC: Spatial Queries In Synchronous Collaboration
Processes of knowledge-building, planning, and decision-making are
frequently accomplished by groups of remotely located people working
together. Research in the field of geocollaboration has been aimed
at discovering ways to support this type of collaborative work using
GIS technologies. One important topic that is yet to be explored is
how to support synchronous collaborative spatial data queries. The
goals of this research are (a) to develop a conceptual framework and
create and assess the Spatial Queries in Synchronous Collaboration
(SQSynC) proof-of-concept prototype to enable remotely located people
to collaboratively query spatial data in real-time requiring only
an internet connection and web browser and (b) to analyze the methods
used by geographers to construct the queries in this environment to
develop a deeper understanding of the process.
The SQSynC software prototype is a web-based application that can
support both simultaneous and parallel collaborative activities between
the users within the interface. In simultaneous mode, the software
behaves as if there is only one interface that every user is concurrently
interacting with, even though each user is remotely connected through
individual computers. For instance, if one person were to pan and
zoom the map, every other user would see the map change. In parallel
mode, the software provides each user with a distinct, private interface
that no other users can see or interact with. To enable more direct
collaboration in the parallel mode, a second, public tab exists. Users
can move items from their private tab over to the public tab in order
to share them with the other users.
A user study was conducted, using the SQSynC prototype, to obtain
targeted input from geographers regarding the two different approaches
to collaborative spatial data query. Participants, working in counter-balanced
pairs, were asked to collaboratively work through one set of tasks
using the simultaneous application mode, then another set of tasks
using the parallel mode or the reverse. In the simultaneous mode,
each user could simultaneously interact with a shared query building
interface. In the parallel mode users could communicate with each
other and share data on the public tab, but there was no shared query
building interface with which they could simultaneously interact,
only a private interface. A total of 16 participants took part in
the experiment and they each completed a follow-up survey to answer
questions about which mode they preferred, report any critical incidents
that occurred or features that they liked, and provide additional
comments and suggestions. The main finding, along with all of the
in-depth feedback from the participants, is that 56% of participants
preferred the parallel mode of collaborative query building, while
44% preferred the simultaneous mode. The results of this experiment
ultimately show that in geocollaboration software there is a place
for both private and shared query building tools.
Kevin's Full Thesis | Video tutorial of SQSynC prototype used for the User Study
More about Kevin's work on CrimeViz
CrimeViz is a web-based map application for exploring spatio-temporal patterns of violent crimes. The application automatically retrieves updated crime data (CRON + RSS) and stores them in a spatial database (PostgreSQL + PostGIS). The client, developed in Flash + AS3, reads in the up-to-date crime data that can then be displayed geographically (Google Maps API) and animated using different temporal aggregations, as set by the user. Kevin worked with Alan MacEachren and Robert Roth to develop the proof-of-concept application in the District of Columbia. Kevin designed a poster along with Craig McCabe and Robert Roth and presented it at the 2009 DHS Summit Student Poster Competition where it was awarded 3rd place (125+ initial submissions). Kevin and Robert also put together a tutorial and code library for creating animated map mashups using the Google Maps API for Flash, which was published in a special digital issue of Cartographic Perspectives. Currently CrimeViz is being transitioned to be used by officers at Harrisburg Police Department and Kevin continues to work on the project remotely.
Poster about CrimeViz | CrimeViz video demonstration | CrimeViz website
