McAlester News-Capital, McAlester, OK

October 4, 2013

OU Spatial Analysis group develops tools to monitor Oklahoma offenders

By Katherine Parker
The Norman Transcript

NORMAN — Helping Oklahoma offenders transition from inmates to productive members of society is a challenge the University of Oklahoma Center for Spatial Analysis was up for while developing a suite of tools to monitor Oklahoma parolees.

With funding provided by the National Institute of Justice, OU partnered with the Oklahoma Department of Corrections to create a web-based tracking analysis application, TRACKS, that will offer more than GPS points and help parole officers in the supervision of offenders.

The OU team consists of May Yuan, professor in the OU College of Atmospheric and Geographic Sciences; Atsushi Nara, geospatial scientist; Marguerite Keesee, sociologist; William Greenwood, IT analyst and web developer; Meredith Denney, graduate research assistant; and Grant Floyd, graduate research assistant.

Location-based offender monitoring systems are widely adopted in Oklahoma and other states as a cost-saving alternative to incarceration for low-risk offenders.

Yuan said the system the Center for Spatial Analysis is developing is different from existing monitoring systems in that it should provide advanced analysis of parolees’ movements and track spatial and temporal patterns.

“The ultimate goal of the project is to help parole officers be more effective,” Yuan said. “Different officers may have different case loads, and watching GPS points on the screen isn’t their job.

“They are trying to get offenders to successfully complete the rehab process. The more we can do to help the parole officers do that, the better.”

TRACKS will provide paroles officers with a variety of tools to monitor behavior, including daily offender movement, geographical features, a social network of potential parolee interaction, hot spots of crime, crime scene correlation and daily alert reports.

Daily offender movement may allow parole officers to determine if a parolee has deviated from normal behavior and violated a term of their parole.

Yuan said that a parolee may stop at a liquor store or a sex offender may stop near a school; TRACKS would present this information to a parole officer.

“Additionally, the geographical features themselves should be helpful because even if an offender doesn’t stop at a liquor store, the parole officer will know if they are near a liquor store and if they are near that location frequently. We hope the system provides officers with information, so they can decide if certain offender activity is worthy of concern,” Yuan said.

Another tool, a social network of potential interaction, will show parole officers the proximate space and time parolees under their watch are likely to have an interaction that would be a violation of their parole.

Hot spots of crime will use crime incident reports from the past two to three years to map out areas with a high density of crime in jurisdictions. Yuan said the team had created such hot spot maps for Tulsa and Norman and are currently working on Oklahoma City.

“Beyond knowing if an offender is in a hot spot, parole officers have to make home visits, and this tool can give them an idea of whether they’re going into a high crime area, so they can prepare,” Yuan said.

Crime scene correlation will provide parole officers data showing their parolees’ GPS points in relation to recent crimes.

Daily alert reports, which Yuan said the team was in the process of creating, should provide parole officers with a compilation of their parolees’ daily parole offenses. For example, a parolee could go to a liquor store and not make curfew; the daily alert report would notify the parole officer of both offenses.

Moreover, Yuan said the team hoped a historical report could be developed to show long-term patterns, not just daily behavior, so parolee behavior might be changed and recidivism decreased.

Currently, in their second user test, Yuan said the team had received helpful feedback from parole officers.

“Parole officers and feedback from the NIJ Correctional Technology Working Group have been very helpful throughout our testing. We hope to have everything worked out by our third user test,” she said.

Yuan said some of the feedback and adjustments the team has had to make has come down to simple visual presentation. Parole officers want to see certain date in a table, not a graph, she said.

The team hopes to have a prototype by late next year, Yuan said.

At the end of the project, OU will make the tracking analysis application publicly available. Corrections and law enforcement organizations interested in TRACKS will be able to download OU’s tools from a password-protected website.

For more information, contact Yuan at myuan@

ou.edu.

Katherine Parker

366-3541

kparker@normantranscript.com