2/5/2025 | 4:10 PM - 5:10 PM | Harbour A

OSINT investigation of social media for classifying reactions on cannabis legalization

Author(s)

Berfin Atabay | SRH Berlin University of Applied Sciences
Resul Gümüs | SRH Berlin
Klaus Schwarz | SRH Berlin University of Applied Sciences
Reiner Creutzburg | TH Brandenburg

Abstract

Cannabis legalization is an important socio-political issue worldwide, creating a variety of public opinions and sentiments. This study uses Open Source Intelligence (OSINT) techniques and methods to analyze social media opinion around the recent legalization of Cannabis in Germany on April 1, 2024, specifically on the Reddit platform. By collecting and analyzing comments from Reddit, we aim to group and understand public sentiment on this topic. Our methodology includes data collection, cleaning, clustering, and sentiment analysis using natural language processing techniques. We use a sentence transformer model for text embedding and the K-means clustering algorithm to find different angles around legalization. Sentiment analysis was implemented using a pre-trained model developed by Hugging Face based on BERT. The results reveal seven distinct clusters representing different themes and sentiments about cannabis legalization. Overall, the sentiment and public opinion tend to lean positive, indicating general public support for Cannabis legalization despite some expectations. In addition to demonstrating the value of OSINT in social media research, this study offers insight into public sentiment surrounding Cannabis legalization to researchers and policymakers.