Using Computational Social Science to Assess Optimal Methods for HIV Prevention Information Diffusion

Background and Purpose: HIV prevention research within the fields of social work and public health has demonstrated the benefit of peer leaders to effect behavior change, through information sharing among high-risk groups. The application of computational models that utilize advances in network analysis and simulation has proven to be cost-effective when there are limited resources for peer leader training. However, there is little research that uses computational modeling to find the optimal selection of peer leaders from communication networks. This presentation will discuss (a) how social network data has been used to assess optimal methods for prevention information diffusion within a population of young Black women (YBW), (b) collaborations with computer scientist and use of artificial intelligence technologies to assess communication networks for HIV intervention development, and (c) other projects currently using computational modeling techniques for HIV intervention development. 

Methods: Respondent driven sampling was used to collect sexual health communication network data from 73 YBW residing in Los Angeles County. Five modes of communication were examined in this study (i.e. in-person, text, phone calls, social media, and a combination of the 4 modes) to determine which method of communication was optimal for disseminating sexual health information through a network of YBW. The “Greedy” algorithm (a maximum influence approximation algorithm) was used to simulate information diffusion within these networks based on (a) mode of communication and (b) selection of potential peer leaders.

Results: Results revealed that a multimodal communication network would yield the most optimal dissemination of sexual health information within this social network of YBW, with 45.73% of YBW being within reach (n=73). Text messaging yielding similar results (41.55%), and social media, in-person, and phone-based networks yielded decent results, indirectly influencing 33-39% of all connections. When examining YBW who already discuss sexual health via texting with network-members (n=64), approximately 45% of YBW could be reached in the network.

 Conclusion and Implementation: These results highlight the variation in the ability to share information within a network depending on the mode of communication used. Few researchers currently use computational modeling techniques in social network-based HIV interventions to optimize information spread throughout their target networks (Eric et al., 2018), with most social network-based interventions designed with limited consideration of the best mode of communication for effective and efficient diffusion of information within targeted social networks. To maximize the diffusion of sexual health and HIV prevention information, researchers who aim to develop peer-led social network-based intervention studies should consider working with computer scientist to help assess what modes of communication are being used by their target populations, how use of the various modes of communication affect information spread within the networks, and ways to better address any network gaps that may impact the effectiveness of social network-based HIV prevention. Knowledge of the communication modalities and their prevalence within the target network is critical in developing culturally tailored social-network based HIV intervention.