CFAR pilot awardee, Dr. Carlos Gallo is an ISGMH affiliate faculty member conducting research that aims to facilitate the implementation evidence-based programs (EBPs) by local agencies, particularly those involved in reducing health inequities in the LGBT and ethnic minority’s populations. He is interested in developing computational methods that monitor and provide real-time feedback of implementation indicators useful for health care providers and funding agencies. He has successfully applied his knowledge in system engineering and computational linguistics in parent-training preventive interventions such as Familias Unidas and New Beginning Programs, funded by NIDA. His work has enhanced the delivery of programs that improves outcomes of Hispanic youth on risky sexual behavior, HIV, and drug abuse. He developed the first machine-based methods that recognize linguistic patterns that are evidence of therapeutic alliance between therapist and family during Familias Unidas home visit sessions. These linguistic patterns are linked to fidelity of implementation and are evidence high fidelity to EBPs protocol. He also developed signal engineering methods to automatically recognize emotion in spoken speech during a EBP delivery in recorded audio/video tapes. Important aspects of the EBPs delivery require providers to speak in specific manners that is either neutral or matches the clients emotions. His work sets the stage for efficient methods of measuring implementation and monitoring and feedback systems that closes the gap between research and practice, during the translation of EBPs in real world use.
Adolescent men who have sex with men (AMSM) experience a disproportionate burden of new HIV diagnoses among all young people. Fortunately, carefully designed mHealth interventions exist to reach and engage this key population. Mobile health (mHealth) is a general strategy to use mobile phones and other wireless mental health interventions. However, these interventions often send scripted messages while ignoring the linguistic style of participants or the linguistic context in which the scripted messages are received. Linguistic style and context affect people’s interpersonal satisfaction and engagement, as demonstrated in sociology, couple’s counseling, and psycholinguistics. For instance, married couples with similar linguistic styles report higher marital satisfaction and are less likely to separate. HIV research has largely ignored how mHealth participants’ linguistic style affects engagement and satisfaction to the intervention.
This lecture will describe computational linguistic methods that analyze the linguistic style of AMSM in order to optimize peer-to-peer platforms of HIV prevention programs. Also, these methods can inform ways to tailor scripted messages to the linguistic context of the peer-to-peer conversation in an efficient, scalable, non-obtrusive, and automatic manner. In summary, this lecture will demonstrate examples where computational linguistic methods could improve the implementation of future generation mHealth HIV interventions.
ISGMH is cosponsoring this event with the Northwestern Department of Linguistics.