Online Behavior Project

This is a PhD dissertation which looks into an automated method for collecting, analyzing, and reporting online sexual solicitations and self reported profiles as an informational component of sexually transmitted infection (STI) management  health programs. It can easily be applied to other health information extraction problems (see Hands on NLP)

AIM 1
Quantitatively characterize known risk factors, other behaviors of interest, and demographics in a population sample of online solicitations. This is looking at a week's worth of activity on two very popular websites in the Seattle area. One of the results of this aim is a typology of online sex seeking behaviors.
.
 
AIM 2
 Identify and evaluate the information needs of researchers and public health officials related to online health-related behavior, particularly concerning STI/HIV. This is done using focus groups, interviews, and online surveys. Some of the results of this work are 'value of information' estimates to help evaluate the output of the computer system.

 

AIM 3  
Develop and evaluate a software tool designed to automatically retrieve, characterize and classify online behavior.

 

Happenings

 

All of the data is in and has been analyzed. I'm currently drafting the dissertation document.

 



Software

There are two bits of custom software for this project. 1) the main application which is implemented in JBoss Seam framework.  And 2) a Firefox plugin that supports it.  Details of these and where you can download the code is coming soon.