Alcohol prevention: a review using Topic Modelling

Background: The science on aetiology of alcohol disorders and effectiveness of preventive interventions for problems related to alcohol use are fragmented and scattered. Research projects have each specific scopes and research questions related to determinants and outcomes measured and underlying theoretical foundations. It is therefore difficult to determine which preventive methods are most effective and which preventive methods remain under researched.

Methods: A comprehensive literature review is carried out, in order to identify the existing evidence concerning alcohol use and its determinants. However, as a traditional literature review is a time-consuming and labour intensive process, often limiting the researchers scope, artificial intelligence is used. Adding more difficulties to the traditional process is the rapidly growing body of knowledge, hindering the researcher to capture an adequate and exhaustive overview. By using artificial intelligence, and more specifically the computational method of topic modelling, these issues can be tackled and provide a more complete overview of the existing body of evidence. Topic modelling is a text-mining technique often used in machine learning and natural language processing. It is an effective method of analysing and summarizing large amounts of textual data, without human intervention, facilitating the identification of hidden topics. 

Aim: This review aims to identify core topics and time trends in research on the prevention of (adolescent) alcohol use through topic modelling. By identifying these core topics, researchers and practitioners get an overview of what is already there and what is missing within the body of evidence. In this way, more targeted research and prevention strategies might be developed.