Alcohol, media, and emerging technologies

This program of work aims to use new and emerging technologies and cutting-edge methodologies to 1) better understand how alcohol is presented in popular media (e.g., social media, movies, music), 2) determine what impact exposure to alcohol-related content can have, and 3) identify what opportunities there are for intervention or prevention.

Specifically, we aim to:

  1. Understand how prevalent alcohol references are in popular media, how alcohol is portrayed (e.g., positive, negative), and how people discuss alcohol-related policies online.
  2. Understand what impact exposure to alcohol in popular media can have on our alcohol use or alcohol-related cognitions.
  3. Identify avenues for intervention, prevention, and policy change and to test and develop tools and interventions that use media platforms or aim to reduce exposure.

Research Team:

Benjamin Riordan, Emmanuel Kuntsche, Dan Anderson-Luxford, Erin Santamaria, Robin Room, Maree Patsouras, Gedefaw Diress, Samatha Salim, Jingda Du, Zhen He (affiliate from Computer Science, La Trobe University), Jennifer Merrill (honorary research fellow from Brown University, US), Taylor Winter (honorary research fellow from the University of Canterbury, NZ).


Have questions regarding our research into alcohol, media, and emerging technology?

Email Benjamin Riordan (

Projects within this area:

Solutions for regulating and monitoring alcohol marketing in Instagram influencer posts

In this study, we aim to determine how common alcohol and alcohol marketing are on Instagram from the top influencers in Australia. Specifically, we aim to answer the questions about: What is the extent of marketing in alcohol-related posts of the most popular Instagram influencers in Australia? What marketing techniques are used in declared and suspected undeclared alcohol marketing practices? Are alcohol marketing posts compliant with law, and what regulatory reforms, if any, are needed to address alcohol influencer marketing? Can a deep learning algorithm identify marketing?

La Trobe Researchers: Emmanuel Kuntsche, Zhen He, Benjamin Riordan, Robin Room, Dan Anderson-Luxford

Funder: VicHealth

Using emerging technologies to estimate the prevalence and impact of digital alcohol exposure

This project aims to use artificial intelligence to quantify the amount of alcohol people are exposed to in digital media (e.g., social media, streaming videos) in their daily lives and the effect alcohol exposure has on alcohol use. Expected outcomes for this project include a quantification of the amount of alcohol exposure in digital media and the impact it has on drinking and a development of a protocol to test exposure. Significant benefits are expected for policy makers aiming to reduce exposure and the public wanting to avoid exposure to limit the harm of alcohol.

Funder: Australian Research Council DE230100659

La Trobe Researchers: Benjamin Riordan, Emmanuel Kuntsche, Zhen He, Dan Anderson-Luxford

External Researchers: Damian Scarf, Rose Marie Ward, Jennifer Merrill, Taylor Winter

A novel social media approach to #identification and #screening for hazardous drinking among diverse non-college young adults

This project is co-led by Associate Professor Jennifer Merrill (Brown University) and Professor Rose Marie Ward (University of Cincinnati) and is funded by National Institute of Health (NIH) in the United States. The project aims to determine content of public social media posts that serves as a marker of hazardous drinking across a diverse group of non-college young adults. To establish acceptability of and obtain feedback on a potential approach to SNS-based screening and outreach for non-college individuals at risk for hazardous drinking.

La Trobe researchers: Benjamin Riordan

External lead researchers: Jennifer Merrill, Rose Marie Ward

Funding: National Institute of Health, R21

Using AI to reveal the true extent & context of alcohol exposure in videos

This project aims to extend an artificial intelligence algorithm to automatically identify and quantify alcohol prevalence in videos. The project is expected to generate significant new knowledge about alcohol’s exposure in these videos’ social, emotional, and environmental contexts. The expected outcomes include a more efficient and automated method of revealing alcohol pervasiveness and its context in the most watched videos in Australia, making costly manual coding redundant. Anticipated benefits include enabling governments to better monitor compliance to alcohol product placement guidelines and increased public awareness of the frequency and harmful effects of being exposed to alcohol in videos.

La Trobe Researchers: Emmanuel Kuntsche, Zhen He, Benjamin Riordan, Aiden Nibali, Erin Santamaria, Samatha Salim

Funding: Australian Research Council DP230100927

Out of sight out of mind? Co-design of a browser plugin to block online alcohol exposure

One simple and scalable solution to reduce the amount of alcohol people are exposed to online is to use a plugin that blocks images with alcohol. Ad blockers are the most popular browser plugin and work by blocking online advertisements. However, the problem is that to reduce normative beliefs and temptations that drive alcohol consumption exposure to alcohol has to be reduced in general, not only alcohol advertisement. Our team have recently developed a deep learning algorithm that can accurately recognise alcohol in images and this technology can be incorporated into a browser plugin to block alcohol imagery, like an ad blocker. However, it is critical that we demonstrate that this approach is acceptable to end users and one key way to improve acceptability is to design the tool with the end users.

Thus, the aim of this project is to use focus groups and semi-structured interviews with drinkers to discuss the acceptability of a browser plugin and to work with them to design the look, feel, and functionality of the plugin.

La Trobe Researchers: Maree Patsouras, Emmanuel Kuntsche, Zhen He, Benjamin Riordan, Amy Pennay, Megan Cook

Funding: School of Psychology and Public Health internal grant

For this project we aim to determine how common alcohol is in music, what is the context of the reference, and to explore whether there is a link between exposure and alcohol use. To answer these questions, we will use a range of methodologies and will conduct systematic reviews and our own content analyses to estimate the pooled prevalence of alcohol references in popular music and music videos and whether these references are increasing over time. We will also develop an algorithm that can be used to determine whether references in music are alcohol-related. Finally, we will use data donations and surveys to determine whether there is a link between alcohol heard daily and alcohol-related cognitions or alcohol use.

La Trobe Researchers: Gedefaw Diress, Dan Anderson-Luxford, Emmanuel Kuntsche, Zhen He, Benjamin Riordan, Aiden Nibali

Funding: La Trobe ABC grant, Australian Research Council DE230100659

The accuracy and promise of zero shot learning

ChatGPT has been labelled a ‘disruptive technology’ due to its ability to answer questions accurately and pass difficult exams. A core reason for the rapid uptake of this technology has been the ability to automatically complete tasks without requiring the user to finetune the model through additional training. The large language model that ChatGPT is built on can do far more than just answer questions and can be used for ‘zero shot learning’ classification tasks to automatically apply labels to previously unlabelled data. ChatGPT and the recent wave of freely available open-source models could ‘change the game’ for researchers who analyse large media data and allow us to understand more about the context of alcohol use in media data rather than simply identify whether it is present or not. We aim to evaluate the performance of large models in classifying alcohol-related text and images compared to human annotators and our specifically trained deep learning algorithms (the current gold standard).

La Trobe Researchers: Emmanuel Kuntsche, Zhen He, Benjamin Riordan, Aiden Nibali, Dan Anderson Luxford

Funding: School of Psychology and Public Health internal grant.