Download This Paper
Preprints with The Lancet is a collaboration between The Lancet Group of journals and SSRN to facilitate the open sharing of preprints for early engagement, community comment, and collaboration. Preprints available here are not Lancet publications or necessarily under review with a Lancet journal. These preprints are early-stage research papers that have not been peer-reviewed. The usual SSRN checks and a Lancet-specific check for appropriateness and transparency have been applied. The findings should not be used for clinical or public health decision-making or presented without highlighting these facts. For more information, please see the FAQs.
Deep Learning and Behavioral Theory: An Improved Analytic Method to Understand HPV Vaccination Intentions from Twitter Discussion
37 Pages Posted: 3 Oct 2019
More...Abstract
Background: HPV vaccine refusal or hesitancy is common among parents of adolescents, which poses a public health threat to communities. An understanding of public perception of the HPV vaccine from the perspective of behavior change theories (e.g., Health Belief Model [HBM], Theory of Planned Behavior [TPB]) can facilitate effective and targeted vaccine promotion strategies. In this regard, social media is considered a major and credible source to access public perception.
Methods: We performed an evaluation of various machine learning and deep learning algorithms in mapping HPV vaccine-related Twitter discussions to major theoretical constructs of HBM and TPB. Deep learning models were then applied to map the sampled Twitter discussions (collected from January 2014 to October 2018) to the theoretical constructs. Locally estimated scatterplot smoothing (LOESS) was then adopted to reveal trends of constructs.
Findings: Deep-learning algorithms achieved better performance compared with machine-learning algorithms on mapping Twitter discussion to the constructs of behavior change theories. The use of pre-trained word embedding can further improve the performance of deep-learning models. LOESS has revealed trends of constructs: for example, in the recent two years Positive attitude toward the HPV vaccine shows an increasing trend, while Negative attitude shows a decreasing trend.
Interpretation: The retrospective analysis of theoretical constructs toward HPV vaccine provides a better understanding of public perception and its evolving trends in terms of multiple dimensions. The increasing prevalence in Positive attitude toward the HPV vaccine benefits from the significant efforts of public health professionals on HPV vaccine promotion; the increase in Perceived severity could result from the promotion strategy that the HPV vaccine has been more associated with cancer prevention than with genital warts.
Funding Statement: Cancer Prevention and Research Institute of Texas, National Institutes of Health, National Science Foundation.
Declaration of Interests: JD received funding from the UTHealth Innovation for Cancer Prevention Research Training Program Predoctoral Fellowship (Cancer Prevention and Research Institute of Texas Grant No. RP160015); JB received funding from the National Science Foundation under Award Number No. 1734134; RS received funding from the CPRIT Dissemination Grant PP190041. CT and YC received funding from the National Institutes of Health under Award Nos. R01LM011829 and R01AI130460.
Ethical Approval Statement: Not required.
Keywords: HPV vaccine, deep learning, Twitter, machine learning, behavioral theory
Suggested Citation: Suggested Citation