Amy M Cohn1,2, Kang Zhao3, Sarah Cha4, Xi Wang5, Michael S Amato4, Jennifer L Pearson6,7, George D Papandonatos8, Amanda L Graham2,4. 1. Battelle Memorial Institute, Arlington, Virginia. 2. Department of Oncology, Georgetown University Medical Center, Washington, DC. 3. Department of Management Sciences, Tippie College of Business, The University of Iowa, Iowa City, Iowa. 4. Schroeder Institute for Tobacco Research and Policy Studies, Truth Initiative, Washington, DC. 5. School of Information, Central University of Finance and Economics, Beijing, China. 6. School of Community Health Sciences, University of Nevada, Reno, Reno, Nevada. 7. Department of Health, Behavior and Society, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland. 8. Center for Statistical Sciences, Brown University School of Public Health, Brown University, Providence, Rhode Island.
Abstract
OBJECTIVE: Alcohol use and problem drinking are associated with smoking relapse and poor smoking-cessation success. User-generated content in online social networks for smoking cessation provides an opportunity to understand the challenges and treatment needs of smokers. This study used machine-learning text classification to identify the prevalence, sentiment, and social network correlates of alcohol-related content in the social network of a large online smoking-cessation program, BecomeAnEX.org. METHOD: Data were analyzed from 814,258 posts (January 2012 to May 2015). Posts containing alcohol keywords were coded via supervised machine-learning text classification for information about the user's personal experience with drinking, whether the user self-identified as a problem drinker or indicated problem drinking, and negative sentiment about drinking in the context of a quit attempt (i.e., alcohol should be avoided during a quit attempt). RESULTS: Less than 1% of posts were related to alcohol, contributed by 13% of users. Roughly a third of alcohol posts described a personal experience with drinking; very few (3%) indicated "problem drinking." The majority (70%) of alcohol posts did not express negative sentiment about drinking alcohol during a quit attempt. Users who did express negative sentiment about drinking were more centrally located within the network compared with those who did not. CONCLUSIONS: Discussion of alcohol was rare, and most posts did not signal the need to quit or abstain from drinking during a quit attempt. Featuring expert information or highlighting discussions that are consistent with treatment guidelines may be important steps to ensure smokers are educated about drinking risks.
OBJECTIVE:Alcohol use and problem drinking are associated with smoking relapse and poor smoking-cessation success. User-generated content in online social networks for smoking cessation provides an opportunity to understand the challenges and treatment needs of smokers. This study used machine-learning text classification to identify the prevalence, sentiment, and social network correlates of alcohol-related content in the social network of a large online smoking-cessation program, BecomeAnEX.org. METHOD: Data were analyzed from 814,258 posts (January 2012 to May 2015). Posts containing alcohol keywords were coded via supervised machine-learning text classification for information about the user's personal experience with drinking, whether the user self-identified as a problem drinker or indicated problem drinking, and negative sentiment about drinking in the context of a quit attempt (i.e., alcohol should be avoided during a quit attempt). RESULTS: Less than 1% of posts were related to alcohol, contributed by 13% of users. Roughly a third of alcohol posts described a personal experience with drinking; very few (3%) indicated "problem drinking." The majority (70%) of alcohol posts did not express negative sentiment about drinking alcohol during a quit attempt. Users who did express negative sentiment about drinking were more centrally located within the network compared with those who did not. CONCLUSIONS: Discussion of alcohol was rare, and most posts did not signal the need to quit or abstain from drinking during a quit attempt. Featuring expert information or highlighting discussions that are consistent with treatment guidelines may be important steps to ensure smokers are educated about drinking risks.
Authors: Ricardo F Muñoz; Alinne Z Barrera; Kevin Delucchi; Carlos Penilla; Leandro D Torres; Eliseo J Pérez-Stable Journal: Nicotine Tob Res Date: 2009-07-29 Impact factor: 4.244
Authors: Kang Zhao; John Yen; Greta Greer; Baojun Qiu; Prasenjit Mitra; Kenneth Portier Journal: J Am Med Inform Assoc Date: 2014-01-21 Impact factor: 4.497
Authors: Thomas M Piasecki; Seungmin Jahng; Phillip K Wood; Brandon M Robertson; Amee J Epler; Nikole J Cronk; John W Rohrbaugh; Andrew C Heath; Saul Shiffman; Kenneth J Sher Journal: J Abnorm Psychol Date: 2011-08
Authors: Robert F Leeman; Sherry A McKee; Benjamin A Toll; Suchitra Krishnan-Sarin; Judith L Cooney; Robert W Makuch; Stephanie S O'Malley Journal: Nicotine Tob Res Date: 2008-12 Impact factor: 4.244
Authors: Katrina A Vickerman; Gillian L Schauer; Ann M Malarcher; Lei Zhang; Paul Mowery; Chelsea M Nash Journal: Biomed Res Int Date: 2015-07-26 Impact factor: 3.411
Authors: Amy M Cohn; Michael S Amato; Kang Zhao; Xi Wang; Sarah Cha; Jennifer L Pearson; George D Papandonatos; Amanda L Graham Journal: Alcohol Clin Exp Res Date: 2018-11-19 Impact factor: 3.455