Presented at the XXXVI Sunbelt Conference of the International Network for Social Network Analysis (INSNA), Newport Beach, CA (April 2016)
During the past two decades, there has been a surge in the number of studies applying social network analysis (SNA) to the study of infectious disease transmission. The mapping of “risk potential networks,” in which individuals are connected by ties that spread infection, has yielded valuable insight into the behavioral epidemiology of human immunodeficiency virus (HIV) and sexually transmitted infections (STI). However, despite these advances and the burgeoning popularity of SNA-based HIV and STI research, methodological challenges remain. SNA’s ability to catalyze major epidemiologic advances relies on researchers’ ability to construct valid representations of participant networks from behavioral data. The standard protocol to construct risk networks, or identify direct and indirect relationships among participants and their partners, involves matching participants’ names and demographic information with the information provided about named partners. This process of identifying and resolving duplicate individuals in the network (known as “entity resolution” (ER)) is often conducted through laborious and time-intensive, manual cross-referencing procedures. In this talk, we will discuss and demonstrate SPIDER, a software package that automatically processes networks of participants and resolves duplicates using both participant characteristics and participant relationships within the network. SPIDER includes a secure server that allows for fine-grained access to research data and high-powered processing of very large networks, and a rich desktop client that allows the research to select and configure a wide range of entity resolution algorithms, reviews the results of entity resolution, and save and share the configuration with other researchers for consistent network design. SPIDER is tailored to the specific needs of HIV/STI researchers (but applies to other domains as well) and will provide them with a system that enables efficient, semi-automated risk network construction using a library of robust, statistically rigorous ER algorithms, rich desktop-based annotation tools, and secure web-based technologies. SPIDER assists researchers in constructing HIV/STI risk networks from data collected on sexual and drug-related partnerships, with the goal of improving methodological quality, comprehensiveness, and standardization.
1 Charles River Analytics
2 University of Kentucky
For More Information
(Please include your name, address, organization, and the paper reference. Requests without this information will not be honored.)