Data Analysis X Social Media Analytics

— Individual Research

Paper: “Analysis of sentiment optimization on social networks based on statistical data”

Project Time Length: Jun 2021 - Sep 2021

Introduction:

Social media has already become an inevitable tool in daily life. Most people believe that social media is convenient for the life of humans since there is no limitation of distance, while people ignore the negative influences of social media. Negative emotions can cause dissatisfaction, depression, and isolation because the virtual companion makes it challenging to overcome the power of face-to-face communication. Meanwhile, people have to digest tons of information on the internet. The research aims to analyze the severity of negative emotions caused by the utilization of social media. The data was collected from the COVID-19 Real World Worry Dataset, selecting emotions like fear, anger, disgust, sadness, and worry from Twitter users as the dependent variable and the duration, the frequency of participate activities on Twitter, the frequency of tweeting, and the frequency for participants on Twitter as the independent variables.

Research Results:

During the data preparation strategy, I identified the expectation of each emotion that a slight level of negative emotion, such as “Worry,” has a smaller influence factor. Based on the results of the linear regression and stepwise selection (bi-direction), I found that the frequency of participants on Twitter is the most significant variable for causing negative emotions. The reason might be the fear of missing comments and invitations from others (9).

Discussions:

The ability to express can impact the study's accuracy based on Pearson’s correlation between the ability to express feelings and negative emotions. Furthermore, I only selected the negative emotions and filtered all excellent emotions, such as happiness, relaxation, and desire, in this study. After years of learning, I noticed that involving these positive emotions in the research is significant since they are also the affective factors for emotions. By including positive emotions, the study can provide higher accuracy and reliability of results. Further research can analyze the severity of negative emotions caused by the use of social media to compare the results before and after the pandemic as well.

Citation：

Chang, J. H., Yuan, Y. X., & Wang, D. (2020). Mental Health Status and Its Influencing Factors among College Students during the Epidemic of COVID-19. Journal of Southern Medical University, 40, 171-176.

Aaron, B. (2021). RWWD full dataset [CSV file]. GitHub.

Robinson, L., & Smith, M. A. (2021). Social media and mental health. HelpGuide.