With every phone swipe and keyboard click, new data is created through tracking searches, sales, and social media conversations. Using machine-based learning and other artificial intelligence-based approaches to rapidly read and analyze this data can allow for monitoring and documenting patterns and trends. While frequently harnessed by business and industry for capturing consumer preferences, the public health community is also turning to social listening and other online data collection.

Recently the National Institute on Drug Abuse of the US Department of Health and Human Services awarded a contract for a research team to develop a “proof of concept machine-learning approach to detect illegal opioid sellers.” While most illegal sales of opioids do not take place online, the government has recognized a need to continue efforts to curtail those sales that do happen. Finding ways to sift through all of the noise to find true transactions is a need – and the hope is that a tool can be developed to help with that.

There have been other efforts to mine social media for sellers of illegal opioids and to help with real-time monitoring of the opioid epidemic. A 2019 study published in the Journal of Medical Internet Research, focused on using deep learning to look for patterns in Instagram posts. In reviewing over 12,000 posts, the researchers found common text in hashtags and comments related to drug trading and identified “1,228 drug dealer posts from 267 unique users.” Another study, published in the Journal of the American Medical Association (JAMA) Network Open, attempted to see whether the application of natural language processing of social media posts could be applied to geospatial data to better track the opioid epidemic. The research team applied machine-learning algorithms to three years worth of Twitter posts geolocated to Pennsylvania. The goal was to see if the machine-learning approach could help in achieving real-time monitoring of the opioid epidemic. By tracking over 200 key words across 16,000 posts, the authors were able to map the Twitter data to the reported state monitoring data.

Successful application of artificial intelligence tools to gather learning on the opioid epidemic and track illegal sales of opioids is just one mechanism to try to keep consumers safe from these dangerous products. CSIP has had a long partnership with many online technology providers who have worked to set up systems to protect consumers. It is a team effort and harnessing the power of machine-learning can only help.

###
About the Center for Safe Internet Pharmacies (CSIP)
The Center for Safe Internet Pharmacies (CSIP), a non-profit organization founded in 2011 by the White House, represents the technology sector and commerce intermediaries including Google, Microsoft, Facebook, Oath, UPS, PayPal, MasterCard, Discover, American Express, and .Health.  CSIP’s mission is to promote industry best practices as it relates to illegal online pharmacies, and educating consumers about safe purchasing of prescription drugs.