Machine Learning Could Help Substance Abuse Stigma

A research team from the University of Waterloo has demonstrated how machine learning (ML) and anonymized data could help address the stigma associated with substance abuse in developing countries, which often makes it difficult to get treatment.

The research paper, titled “A Machine Learning Model for Predicting Individual Substance Abuse with Associated Risk-Factors,” was published in the journal Annals of Data Science.

Insight Into Underlying Factors

The new approach provided insight into the underlying factors that influence substance abuse tendencies. It provides a brand new look into a subject that is often surrounded by social and cultural taboos.

The research identified several significant risk factors, such as family relationships, a curiosity to experiment with drugs, and relationships with friends who also suffer from substance abuse.

Enamul Haque is a PhD researcher in computer science at the University of Waterloo and lead author of the research.

“In a country like Bangladesh, people can be hesitant to discuss substance abuse issues,” Haque said. “This kind of research will enable policy-makers to have better information and then be able to design better programs to help address substance abuse.”

Training ML Algorithms to Identify Risk Factors

The new research was based on data pulled from various sources, such as one-on-one interviews and mass online surveys. The survey data was mostly sourced from developing countries in South Asia.

“Within the countries where we conducted the survey, we collected data from a broad and diverse pool of respondents,” Haque continued. “We looked for different respondents based on age, gender and socio-economic context.”

The team first collected a massive amount of data to be used in the study. They then relied on machine-learning algorithms to identify patterns and key risk factors of substance abuse. In order to carry out the computer science part of the research, the team set up multiple stages of data analysis and refinement.

“I really hope this research can help people dealing with substance abuse issues and get them the support they need,” Haque said.

Co-authors of the research included Uwaise Ibna Islam, Dheyaaldin Alsalman, Muhammad Nazrul Islam, Mohammad Ali Moni, and Iqbal H. Sarker.

This new approach is one of the many examples of how AI and machine learning can be used to address several psychological and physical addictions. These technologies provide many opportunities to develop innovative treatments for the future, as well as to understand the underlying factors contributing to each addiction.