Applying Machine Learning to Identify Anti-Vaccination Tweets during the COVID-19 Pandemic

Anti-vaccination attitudes have been an issue since the development of the first vaccines. The increasing use of social media as a source of health information may contribute to vaccine hesitancy due to anti-vaccination content widely available on social media, including Twitter. Being able to identify anti-vaccination tweets could provide useful information for formulating strategies to reduce anti-vaccination sentiments among different groups. This study aims to evaluate the performance of different natural language processing models to identify anti-vaccination tweets that were published during the COVID-19 pandemic. We compared the performance of the bidirectional encoder representations from transformers (BERT) and the bidirectional long short-term memory networks with pre-trained GLoVe embeddings (Bi-LSTM) with classic machine learning methods including support vector machine (SVM) and naïve Bayes (NB). The results show that performance on the test set of the BERT model was: accuracy = 91.6%, precision = 93.4%, recall = 97.6%, F1 score = 95.5%, and AUC = 84.7%. Bi-LSTM model performance showed: accuracy = 89.8%, precision = 44.0%, recall = 47.2%, F1 score = 45.5%, and AUC = 85.8%. SVM with linear kernel performed at: accuracy = 92.3%, Precision = 19.5%, Recall = 78.6%, F1 score = 31.2%, and AUC = 85.6%. Complement NB demonstrated: accuracy = 88.8%, precision = 23.0%, recall = 32.8%, F1 score = 27.1%, and AUC = 62.7%. In conclusion, the BERT models outperformed the Bi-LSTM, SVM, and NB models in this task. Moreover, the BERT model achieved excellent performance and can be used to identify anti-vaccination tweets in future studies.

Changes in depressive symptoms and correlates in HIV+ people at An Hoa Clinic in Ho Chi Minh City, Vietnam

BACKGROUND

Understanding of depression among Vietnamese people living with HIV (PLWH) is limited. This longitudinal study examines changes in depressive symptoms and identifies its correlates among people living with HIV under antiretroviral therapy at An Hoa Clinic.

METHODS

People living with HIV ≥18 years and undergoing antiretroviral therapy for ≥3 months were eligible. Those at final AIDS stage, too ill, or illiterate were excluded due to their inability to complete the self-administered questionnaire. One researcher was present in the clinic for a month inviting PLWH to participate. Data were collected from 242 PLWH at baseline (T1) and 234 after three months (T2). Depressive symptoms was measured by the Center for Epidemiologic Studies Depression Scale (CESD). Social relationship was measured using questions created by World Health Organization. Generalized Estimating Equations were used examining changes in depressive symptoms with CESD cut-off <16/≥16 (mild depression) and cut-off <23/≥23 (major depression).

RESULTS

Model 1 (CESD cut-off <16/≥16) showed that participants were not more likely to have depressive symptoms at T2 compared to T1 (OR = 1.15, p > 0.05). Those with a co-morbidity were more likely to have depressive symptoms than those without a co-morbidity (OR = 1.76, p < 0.05). Those with higher social relationship scores were less likely to have depressive symptoms than those with lower scores (OR = 0.76, p < 0.001). Model 2 (CESD cut-off <23/≥23) showed that participants were more likely to have major depressive symptoms at T2 compared to T1 (OR = 1.6, p < 0.01) and those with higher social relationship score were less likely to have major depressive symptoms than those with lower scores (OR = 0.73, p < 0.001).

CONCLUSIONS

People living with HIV were not more likely to have depressive symptoms (<16/≥16) but were more likely to have major depressive symptoms (<23/≥23) at T2 vs. T1. Social relationship was found to be strongly associated with depressive symptoms. Associations between age, individual income status, and co-morbidity with depressive symptoms were not decisive. Gender, ethnicity, education, religion, marriage, household economy, and adherence were not correlates.