Document Type


Publication Date


Publication Source

Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM 2019)


In this paper, we focus on the collection and analysis of relevant Twitter data on a state-by-state basis for (i) measuring public opinion on marijuana legalization by mining sentiment in Twitter data and (ii) determining the usage trends for six distinct types of marijuana. We overcome the challenges posed by the informal and ungrammatical nature of tweets to analyze a corpus of 306,835 relevant tweets collected over the four-month period, preceding the November 2015 Ohio Marijuana Legalization ballot and the four months after the election for all states in the US. Our analysis revealed two key insights: (i) the people in states that have legalized recreational marijuana express greater positive sentiments about marijuana than the people in states that have either legalized medicinal marijuana or have not legalized marijuana at all; (ii) the states that have a high percentage of positive sentiment about marijuana is more inclined to authorize (e.g., by allowing medical marijuana) or broaden its legal usage (e.g., by allowing recreational marijuana in addition to medical marijuana). Our analysis shows that social media can provide reliable information and can serve as an alternative to traditional polling of public opinion on drug use and epidemiology research.

Inclusive pages

952 - 961



Document Version

Published Version


This open-access article is provided for download in compliance with the publisher’s policy on self-archiving. To view the version of record, use the DOI:


ASSOC Computing Machinery

Peer Reviewed