WeeDetection: Multi-platform Detection System to Identify and Report Illegal Sales of Cannabis-derived Products

1.a. PROJECT SUMMARY/ABSTRACT The proposed project will address a critical public health challenge: developing technology that can help the cannabis industry, law enforcement, and regulators with addressing the growing black market for cannabis products via online platforms. While many states have now legalized medical and recreational use of cannabis, others continue to strictly regulate or prohibit its use. This has led to confusion and lack of policy coherence, creating opportunities for unscrupulous providers to divert or sell unregulated, adulterated, and otherwise unsafe cannabis products via the Internet. This issue is further complicated by new and emerging cannabis products with unknown and questionable safety that can be marketed and sold directly to consumers on popular online platforms, including social media, e-commerce sites, and the dark web. In response, this project will utilize advances in big data, machine learning, and data visualization to establish a multiplatform detection and compliance system to identify, classify, and report illegal online cannabis sales with the aim of enabling digital prevention and harm reduction. The system will develop a digital database that cross-references state licensure information for compliance checking and identifying specific violating online sellers to prevent customer exposure and associated health harms. This novel approach will enable the development of a robust end-to-end automated cannabis product online monitoring and compliance solution needed to ensure the future viability and integrity of the legitimate U.S. cannabis market. The project has the following project aims: Milestone 1: Establish dedicated data collection protocols for multiple social media channels, major cannabis e-commerce platforms, and dark web marketplaces, filtered for cannabis product and consumption-related keywords (M1: Create a multiplatform data collection system and database for cannabis products and sellers) Milestone 2: Develop sufficient training data and machine learning algorithms to identify and classify illegal cannabis sellers with high precision and efficiency (M2: develop a suite of algorithms and conduct model evaluation) Milestone 3: Build data visualization tools and compliance solutions with specific end user feedback to create a customizable front-end, web hosted data dashboard that enables detection and automated reporting of illegal sellers (M3: Create an MVP based on needs of customer segments and conduct business hypothesis testing) This project will fill existing gaps in illicit cannabis research and innovation through the development of a customizable and cost-efficient tool to improve the identification, detection, and reporting of unregulated and illegal cannabis sales to ensure the integrity of the legal cannabis supply chain and protect consumers.