# Using Machine Learning to Detect and Characterize Synthetic Psychoactive Drug Digital Marketing and Distribution

> **NIH NIH R21** · UNIVERSITY OF CALIFORNIA, SAN DIEGO · 2021 · $197,344

## Abstract

The study will address a critical challenge regarding the public health threat of synthetic psychoactive drugs
(SPDs): identifying and characterizing how social media is used to market, distribute and sell these products to
the public. We will carry out these aims by examining large volumes of data from popular social media
platforms Twitter and Instagram using an innovative methodology involving machine learning and
multidisciplinary data analysis. This research project addresses a key NIDA research objective: “research on
factors which influence initiation, continuation and desistance of use of specific SPDs in different populations.
Examples include:…market and distribution factors; policies; and the role of peer influences, social media and
the internet.” We will first collect large volumes of data filtered for SPD drug diversion, marketing, and drug
dealing via social media posts, comments, and other messages. We will then code the data using an
advanced machine learning protocol. Finally, we will analyze the data using statistical, geospatial and network
analysis to assess associations between SPD marketing and specific risk factors for user groups. We will
accomplish these objectives using a multidisciplinary approach that involves disciplines and methods from
public health-epidemiology, substance abuse research, computation science, legal and policy analysis, and
geospatial analysis. The project has the following specific aims:
 Aim 1: Conduct digital surveillance of popular social media platforms Twitter and Instagram in order to
 describe the nature and magnitude of online marketing, sale and distribution of SPDs;
 Aim 2: Characterize the types of marketing strategies used by SPD vendors and drug dealers on these
 platforms, including marketing messages used to influence knowledge, attitudes and perceptions
 pertaining to SPDs and identification of the users and networks they target;
 Aim 3: Using statistical, network and geospatial methods, describe, test, and visualize associations
 between SPD marketing and distribution and specific risk factors and user groups.
This is a critical opportunity to better understand how social media contributes to a growing “digital” risk
environment that can influence user initiation and enable illegal access to SPDs. Though research examining
the linkages between social media and substance abuse has been growing, no study has specifically
examined online marketing and access characteristics on more than one platform while at the same time
assessing how geographic and user network factors can impact this unique and largely unregulated risk
environment. Results from this study can generate further study hypotheses on the association between SPD
behavioral risk factors and SPD product trends, while also informing future interventions utilizing e-health tools
including targeted regulation, health education, and counter-marketing.

## Key facts

- **NIH application ID:** 10142413
- **Project number:** 5R21DA050689-02
- **Recipient organization:** UNIVERSITY OF CALIFORNIA, SAN DIEGO
- **Principal Investigator:** Timothy Ken Mackey
- **Activity code:** R21 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2021
- **Award amount:** $197,344
- **Award type:** 5
- **Project period:** 2020-04-15 → 2023-03-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10142413

## Citation

> US National Institutes of Health, RePORTER application 10142413, Using Machine Learning to Detect and Characterize Synthetic Psychoactive Drug Digital Marketing and Distribution (5R21DA050689-02). Retrieved via AI Analytics 2026-05-22 from https://api.ai-analytics.org/grant/nih/10142413. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*