A Study On The Prediction Of The Adoption Of Health Information By Social Media Users Under Public Health Emergencies

Objective:Based on the theoretical model of information adoption,the influencing factors of the adoption behavior of social media users under public health emergencies are determined,and a prediction model of health information adoption behavior of social media users under public health emergencies is constructed and evaluated on this basis.It provides a certain reference significance for the selection of predictive feature indicators in future prediction research related to user information behavior in other fields,and also provides important practical guidance for social media to better improve the level of information services in the future.Methods:In this study,a small-scale user interview was conducted under the guidance of the existing theoretical model of information adoption,and a total of 14 interviewees were invited to analyze the interview data through the method of grounded theory,and constructed a model of influencing factors influencing the adoption behavior of social media users’ health information under public health emergencies.According to the influencing factor model,a total of 306 valid questionnaires were collected,and statistical methods such as descriptive statistics,one-way ANOVA,correlation analysis,and regression analysis were used to conduct statistical analysis of the questionnaire data and analyze the degree of influence of each influencing factor on information adoption behavior;After the questionnaire data is preprocessed to form a sample set,the machine learning method of the support vector machine is used to establish a prediction model for the adoption of social media users’ health information under public health emergencies,and the results are analyzed and evaluated.Results:(1)The interview results of 14 interview subjects were analyzed by rooting theory three-level coding,and the three core codes of environmental factors,user factors and information factors were obtained,and a model of influencing factors of social media users’ health information adoption behavior under public health emergencies was constructed according to the three-level coding table.(2)According to the questionnaire designed according to the influencing factor model,the 306 valid questionnaire data collected were statistically analyzed,and the interpersonal impact,health risk cognition,individual status,individual motivation,perceptual expectation,source credibility,information content quality,and information dissemination quality could explain the 67.8% change in information adoption behavior.The regression coefficient value for individual state was 0.066(t=3.066,P=0.002<0.01),the regression coefficient value for health risk perception was 0.156(t=2.179,P=0.025<0.05),the regression coefficient value for perceptual expectation was 0.094(t=1.966,P=0.047<0.05),the regression coefficient value for individual motivation was0.118(t=2.584,P=0.010<0.05),and the regression coefficient value for interpersonal impact was 0.164(t=5.029,P<0.01),the regression coefficient value for information content quality is 0.112(t=2.255,P=0.025<0.05),the regression coefficient value for source confidence is 0.055(t=1.903,P=0.046<0.05),the regression coefficient value for information dissemination quality is 0.104(t=3.618,P<0.01),the regression coefficient value for age is 0.535(t=8.215,P<0.01),and the regression coefficient value for medical knowledge background is-0.042(t=-1.991,P=0.047<0.05),these 10 dimension variables can be used as input features for the prediction model.(3)Using the support vector organization to build a prediction model,the prediction indicators from different angles,namely environmental factors + user factors+ information factors,environmental factors + user factors + information factors +demographic characteristics,information factors as input variables for machine learning,the prediction accuracy of the final model is obtained: 83.79%,83.85%,75.32%,and all >70%.Conclusion:(1)Environmental factors,user factors,and information factors will affect the adoption behavior of social media users’ health information under public health emergencies.(2)Interpersonal impact,individual status,individual motivation,health risk cognition,perceived expectations,source credibility,information content quality,and information dissemination quality all have a significant positive impact on information adoption behavior.(3)When subjective factor demographics(age,medical background)are added as predictors,the prediction model has the highest accuracy.