The UO’s Computer and Information Science department will host a forum June 5 on data-mining, social networks and the impact on health-care outcomes.

Dejing Dou, a professor in the department, and his SMASH research team will host a forum on their project, “Understanding the Mechanism of Social Network Influence in Health Outcomes through Multidimensional and Semantic Data Mining Approaches.”

The forum will be held in the Deschutes Hall Colloquium Room. The morning session, from 9 a.m. to noon, is free and open to the public.

This session will include opening remarks from Kimberly Espy, university vice president for research; Ran Whitehead, CEO of PeaceHealth Laboratories; and technical presentations on PeaceHealth’s YesiWell program by Dr. Brigitte Piniewski and David Kil; the SMASH graph and influence model; the impact of the YesiWell social network on physical activity and weight loss; and on privacy preserving social network analysis.

SMASH stands for Semantic Mining of Activity, Social, and Health data Project. This multi-disciplinary team is a collaboration between UO CIS and partners that include PeaceHealth, Kent State, University of North Carolina-Charlotte and George Washington University.

The team was recently awarded a three-year, $1.5 million grant from the National Institutes of Health/National Institute of General Medical Sciences. The research will address the need for formal ontologies, data mining, graph mining and privacy-preserving tools to help understand the influence of healthcare social networks on sustained weight loss, where the data are multi-dimensional, temporal, semantically heterogeneous and sensitive.

"About 2/3 of adults in America are overweight," Dejing said. "The SMASH project is the first joint NIH project between UO and PeaceHealth Labs. It will show how good health behaviors -- for example, exercise -- spread in social networks and help overweight people lose weight."

The afternoon sessions are by invitation only.  During them, the research team will map out the SMASH research agenda, with focused discussions on the influence model, ontologies and mappings, recommendation and intervention, privacy and the web portal.

- by UO’s Computer and Information Science Department