PEER SURE App

From REU@MU
Revision as of 06:35, 3 June 2023 by Khatami (Talk | contribs)

Jump to: navigation, search

Project Background

Substance use disorder (SUD) is a major public health problem that affects millions of people worldwide. One of the promising approaches for SUD prevention and treatment is peer support, which involves people who have lived experience of SUD or recovery providing information, advice, and emotional support to others who are facing similar challenges. However, peer support for SUD faces some barriers and limitations, such as lack of accessibility, availability, affordability, and diversity of peer support services and resources. Therefore, there is a need for innovative and flexible ways to deliver peer support for SUD that can overcome these challenges and enhance the peer support experience.

One of the potential solutions is to use web-based technologies to facilitate peer support for SUD. Web-based technologies can offer several advantages for peer support for SUD, such as increasing the reach, convenience, anonymity, and personalization of peer support services and resources. Web-based technologies can also enable new forms of peer interaction and engagement that can foster a sense of community and belonging among people with SUD.

One of the emerging web-based technologies that can be applied to peer support for SUD is generative artificial intelligence (AI), which is a system that can generate natural language text based on some input or context. Generative AI can potentially create realistic and personalized texts for people with SUD based on their input and preferences, such as stories, poems, or motivational messages. Generative AI can also potentially provide feedback and guidance to people with SUD based on their progress and performance.

However, generative AI also poses some challenges and risks for peer support for SUD, such as the quality and accuracy of the generated texts, the ethical and legal implications of using generative AI to provide SUD advice or support, and the emotional and social aspects of peer interaction with generative AI. Therefore, there is a need for careful design, implementation, and evaluation of generative AI for peer support for SUD to ensure its quality, accuracy, ethics, and effectiveness.

Project Description

The aim of my part in this project is to search how to implement an AI plug in and figure how it would work with the main web-based application that uses generative AI to enhance peer support for SUD. I will conduct a literature review on generative AI for SUD to identify the best practices and the gaps in the field. I will use existing tools and APIs to train a large language model on the peer mentoring texts for SUD and evaluate its quality and accuracy using various metrics and methods. I will test and refine the large language model using user feedback and expert input. I will integrate the large language model with the main web-based application that allows users to interact with generative AI texts that are tailored to their needs and preferences regarding SUD prevention or treatment.

Weekly Goals

Week Description
Week 1
  • Meeting the cohort and advisors
  • Attending Orientation session
  • Attending Good Research Practice session
  • Conducting a literature review on generative AI for SUD to identify the best practices and the gaps in the field
  • Opening a page in Marquette Wiki for Project milestones and goals |