Community-Driven Data Engineering for Substance Abuse Prevention in the Rural Midwest
|Community-Driven Data Engineering for Substance Abuse Prevention in the Rural Midwest|
|Motto||To help small and rural communities in the Midwest address the opioid epidemic via BIGDATA (BD) technology|
|Timeline||September 1, 2018 - November 30, 2019|
|Funding Agency||National Science Foundation|
|Award Number||OAC 1761931|
BD Spokes: Community-Driven Data Engineering for Substance Abuse Prevention in the Rural Midwest is a NSF funded project involving a collaboration between AI Institute, University of South Carolina and Ohio State University.
The opioid crisis ravaging Ohio and the Midwest disproportionally affects small and rural communities. Harnessing and deploying data holds promise for developing a response to this crisis by policymakers, healthcare providers, and citizens of the communities. Currently, there are many barriers to getting data into the hands of individuals on the frontlines. Crucial data are siloed across law enforcement, public health departments, hospitals and clinics, and county administrations; data often are inaccurate or collected in non-standard ways across different agencies and departments; the stigma of drug abuse limits accurate reporting of drug-related deaths; and information is not shared with the community and other stakeholders because of the lack of a privacy and security framework. Such barriers, for example, prevent individuals with addictions or their families and friends from locating available treatment centers or obtaining other important information in a timely way. Similarly, it is difficult for first responders and healthcare providers to obtain critical up-to-date information. In predominantly rural counties, these challenges are especially daunting because there is often poor connectivity and communication infrastructure. This Big Data Spoke project involves developing scalable, flexible, and connectivity-rich data-driven approaches to address the opioid epidemic. The cyberinfrastructure framework, OpenOD, will be initially designed and deployed in small and rural communities in Appalachia Ohio and the Midwest, where the need for data and connection are greatest. Based upon significant community input, OpenOD will also create end-user applications or enterprise solutions to support stakeholders and communities to mount a response they feel will be most efficient and beneficial at the local level. As a Spoke to NSF?s Midwest Big Data Hub, our efforts can be efficiently scaled, disseminated, and applied to the opioid and other societal problems such as infant mortality, crime, and natural disasters. This project fits within NSF's mission to promote the progress of science (contribute to the science and engineering of large socially relevant cyberinfrastructures) to advance the health and welfare of US citizens (by linking data sources in new and useful ways to empower communities to address societal problems; establishing sustainable partnerships between academia, industry, government and communities; increasing data literacy and community engagement with data science; and enhancing research and education via development/adaptation of training modules and courses in data analytics).
The main goal of this project is to help small and rural communities in the Midwest address the opioid epidemic via BIGDATA (BD) technology. While no communities have been spared, small and rural communities face unique challenges in confronting the opioid epidemic: knowledge and data exist in siloes across multiple organizations with varying jurisdictional boundaries; efforts to collect, link, and analyze data are hampered by a lack of infrastructure and tools; rural areas are plagued by "dead zones" in cellular connectivity; communities lack capacity for data collection, and analytics; needs and resources across effected communities are not uniform and require BD approaches that are flexible, open, leverage significant community input, and can be dutifully validated. Our proposed solution is OpenOD, a framework that provides uniform, relevant and timely access to data. Working integrally with the Midwest Big Data Hub (MBDH) and our partners, our three main objectives are to: (1) Work with local communities to understand strengths and gaps in cyberinfrastructure, data availability, and need for data analytics workforce skills. (2) Assemble flexible cyberinfrastructure that includes a data commons, stakeholder-usable and cloud-amenable data analytics and visualization tools, and internet connectivity with both mobile and non-mobile capabilities. (3) Validate, evaluate, and disseminate cyberinfrastructure and data analytics tools to stakeholder groups throughout the region while fostering new partnerships. OpenOD will create approaches that will allow governing units to deploy openly available tools rather than rely on proprietary tools. In this way, existing disparities in data access and ensuing responses are effectively addressed. The potential contributions of the project are to: (1) Increase BD and STEM literacy and community engagement in underrepresented groups given the operating milieu of OpenOD in rural areas where the population is indigent and lacks adequate skills to join the modern workforce. (2) Improve well-being of individuals in society by linking data sources in new and useful ways to empower communities to address the opioid crisis; improved connectivity and timely delivery of critical information will accelerate community responsiveness and improve preventive strategies. (3) Provide infrastructure for research and education will be improved given that project activities will deliver linked, curated data sets to community stakeholders, researchers and educators. Training modules and courses adapted and developed and shared with local/regional educators and will remain with the communities after the funding period has ended. In addition, new and established partnerships will allow sustainability of the project in the communities for the long-term.
Principal Investigators: Prof. Amit P. Sheth (AIISC, UofSC)
Postdoctoral Researchers: Dr. Ugur Kursuncu
- Lamy, Francois R., Raminta Daniulaityte, Monica J. Barratt, Usha Lokala, Amit Sheth, and Robert G. Carlson. "Listed for sale: analyzing data on fentanyl, fentanyl analogs and other novel synthetic opioids on one cryptomarket." Drug and Alcohol Dependence (2020): 108115.
- Kumar, Ramnath, Shweta Yadav, Raminta Daniulaityte, Francois Lamy, Krishnaprasad Thirunarayan, Usha Lokala, and Amit Sheth. "eDarkFind: Unsupervised Multi-view Learning for Sybil Account Detection." In Proceedings of The Web Conference 2020, pp. 1955-1965. 2020.
- Usha Lokala, Francois R. Lamy, Raminta Daniulaityte, Amit Sheth, Ramzi W. Nahhas, Jason I. Roden, Shweta Yadav, and Robert G. Carlson. "Global trends, local harms: availability of fentanyl-type drugs on the dark web and accidental overdoses in Ohio." Computational and Mathematical Organization Theory 25, no. 1 (2019): 48-59.
- Ugur Kursuncu, Manas Gaur, Usha Lokala, Krishnaprasad Thirunarayan, Amit Sheth and I. Budak Arpinar. "Predictive Analysis on Twitter: Techniques and Applications". Book Chapter in "Emerging Research Challenges and Opportunities in Computational Social Network Analysis and Mining", Editor: Nitin Agarwal, Springer, 2019.
- Ugur Kursuncu, Manas Gaur, Usha Lokala, Anurag Illendula, Krishnaprasad Thirunarayan, Raminta Daniulaityte, Amit Sheth, and I. Budak Arpinar. "What's ur Type? Contextualized Classification of User Types in Marijuana-Related Communications Using Compositional Multiview Embedding." In 2018 IEEE/WIC/ACM International Conference on Web Intelligence (WI), pp. 474-479. IEEE, 2018.
- Gaur, M., Kursuncu, U., Sheth, A. Yadav, S. & Wickramarachchi (2020), "Hypertext 2020 Tutorial: Knowledge-infused Deep Learning", In 31st ACM Conference on Hypertext and Social Media (HT'20), Florida, USA