Self-presentation In Social Networks And Subjective Well-bein

Self-presentation refers to an individual’s efforts in social interaction to present oneself in a manner that shapes others’ perceptions in accordance with one’s own desires.Previous research has revealed that positive self-presentation can significantly enhance an individual’s subjective well-being.However,there is limited attention to the negative consequences of positive self-presentation on individuals within social networks.The success proposition of social interaction highlights that the more frequently an individual’s behavior is rewarded,the more likely they are to repeat that behavior.In the context of self-presentation,online feedback incentivizes individuals to increase their frequency of behavior on social networks,potentially leading to social network addiction and a subsequent decrease in subjective well-being.Social interaction is one of the key factors influencing subjective well-being.Previous studies that predict subjective well-being have primarily focused on real-life behavior,neglecting modeling in virtual network contexts.This study builds on social interaction theory and employs machine learning modeling techniques to predict individuals’ subjective well-being based on their self-reported data(positive selfpresentation,true self-presentation,online positive feedback,social self-efficacy,social network addiction,social comparison)and social network behavior characteristics on "Sina Weibo".Study 1,explored the relationship between positive self-presentation and subjective well-being in social networks,and the mechanism of social network addiction in this relationship.232 college students were surveyed by positive selfpresentation,subjective well-being and social network Addiction Tendency Scale.The results showed that:(1)social network positive self-presentation can positively predict subjective well-being;(2)Social network addiction mediates the relationship between positive self-presentation and subjective well-being,and shows the masking effect.Specifically,positive self-presentation positively affects social network addiction,while social network addiction negatively affects subjective well-being.The conclusion shows that social network addiction can suppress the influence of positive selfpresentation on subjective well-being.Study 2,social interaction is one of the important influencing factors of subjective well-being.To explore the impact of self-presentation and other potential factors on subject well-being in the process of social interaction.This study collected self-report questionnaire data from 316 participants and established a machine learning model.The questionnaire included self-presentation,online positive feedback,social self-efficacy,social network addition,social comparison,and subjective well-being.The results showed that machine learning models based on social interaction achieved good predictive performance.Among them,the XGBoost model had an accuracy of 85.72%and an F1 score of 0.858 in predicting subjective well-being,which was higher than the other algorithms.The conclusion indicates that positive self-presentation,authentic self-presentation,online positive feedback,social self-efficacy,social network dependency,and social comparison in social network contexts can significantly predict individuals’ subjective well-being.Study 3,To enrich the machine learning model of subjective well-being,this study added individual social network behavior features on the self-report data used in Study2.This study collected self-reported questionnaire data(same as Study 2)from a total of 230 participants(65 males,165 females;age 23.11 ± 3.78 years)and used web crawling techniques to retrieve social network behavioral characteristics data based on the Weibo IDs provided by the participants.The combination of the two sources of data formed a multimodal dataset,which was fitted and trained using decision tree models,random forest models,XGBoost models,and Ada Boost models.The results showed that all four machine learning models could effectively predict subjective well-being,and the prediction performance of the multimodal model was superior to that of the single-modal machine learning models for each model.Among them,the random forest model based on multimodal data had the best prediction performance,with an accuracy of 84.82%,an F1 score of 0.830,and an area under the ROC curve of 0.805.The conclusion suggests that the prediction performance of a multimodal machine learning model that unifies subjective self-reported data with objective behavioral data is significantly improved.In summary,the relationship between positive self-presentation on social networks and subjective well-being is inhibited by social network dependence.Moreover,Combining subjective self-reported data with objective behavioral data can also enhance the model’s performance.