Language sentiment predicts changes in depressive symptoms.

The prevalence of depression is a major societal health concern, and there is an ongoing need to develop tools that predict who will become depressed. Past research suggests that depression changes the language we use, but it is unclear whether language is predictive of worsening symptoms. Here, we test whether the sentiment of brief written linguistic responses predicts changes in depression. Across two studies (N = 467), participants provided responses to neutral open-ended questions, narrating aspects of their lives relevant to depression (e.g., mood, motivation, sleep). Participants also completed the Patient Health Questionnaire (PHQ-9) to assess depressive symptoms and a risky decision-making task with periodic measurements of momentary happiness to quantify mood dynamics. The sentiment of written responses was evaluated by human raters (N = 470), Large Language Models (LLMs; ChatGPT 3.5 and 4.0), and the Linguistic Inquiry and Word Count (LIWC) tool. We found that language sentiment evaluated by human raters and LLMs, but not LIWC, predicted changes in depressive symptoms at a three-week follow-up. Using computational modeling, we found that language sentiment was associated with current mood, but language sentiment predicted symptom changes even after controlling for current mood. In summary, we demonstrate a scalable tool that combines brief written responses with sentiment analysis by AI tools that matches human performance in the prediction of future psychiatric symptoms.

Determining containment policy impacts on public sentiment during the pandemic using social media data.

Stringent containment and closure policies have been widely implemented by governments to prevent the transmission of COVID-19. Yet, such policies have significant impacts on people's emotions and mental well-being. Here, we study the effects of pandemic containment policies on public sentiment in Singapore. We computed daily sentiment values scaled from −1 to 1, using high-frequency data of ∼240,000 posts from highly followed public Facebook groups during January to November 2020. The lockdown in April saw a 0.1 unit rise in daily average sentiment, followed by a 0.2 unit increase with partially lifting of lockdown in June, and a 0.15 unit fall after further easing of restrictions in August. Regarding the impacts of specific containment measures, a 0.13 unit fall in sentiment was associated with travel restrictions, whereas a 0.18 unit rise was related to introducing a facial covering policy at the start of the pandemic. A 0.15 unit fall in sentiment was linked to restrictions on public events, post lock-down. Virus infection, wearing masks, salary, and jobs were the chief concerns found in the posts. A 2 unit increase in these concerns occurred even when some restrictions were eased in August 2020. During pandemics, monitoring public sentiment and concerns through social media supports policymakers in multiple ways. First, the method given here is a near real-time scalable solution to study policy impacts. Second, it aids in data-driven and evidence-based revision of existing policies and implementation of similar policies in the future. Third, it identifies public concerns following policy changes, addressing which can increase trust in governments and improve public sentiment.

COVID-19, climate change, and the finite pool of worry in 2019 to 2021 Twitter discussions.

Climate change mitigation has been one of the world's most salient issues for the past three decades. However, global policy attention has been partially diverted to address the COVID-19 pandemic for the past 2 y. Here, we explore the impact of the pandemic on the frequency and content of climate change discussions on Twitter for the period of 2019 to 2021. Consistent with the "finite pool of worry" hypothesis both at the annual level and on a daily basis, a larger number of COVID-19 cases and deaths is associated with a smaller number of "climate change" tweets. Climate change discussion on Twitter decreased, despite 1) a larger Twitter daily active usage in 2020 and 2021, 2) greater coverage of climate change in the traditional media in 2021, 3) a larger number of North Atlantic Ocean hurricanes, and 4) a larger wildland fires area in the United States in 2020 and 2021. Further evidence supporting the finite pool of worry is the significant relationship between daily COVID-19 cases/deaths on the one hand and the public sentiment and emotional content of climate change tweets on the other. In particular, increasing COVID-19 numbers decrease negative sentiment in climate change tweets and the emotions related to worry and anxiety, such as fear and anger.

Cost-effectiveness of 4CMenB vaccination against gonorrhea: importance of dosing schedule, vaccine sentiment, targeting strategy, and duration of protection.

BACKGROUND: Observational evidence suggests the 4CMenB meningococcal vaccine may partially protect against gonorrhea, with one dose being two-thirds as protective as two. We examined the cost-effectiveness of vaccinating men-who-have-sex-with-men (MSM) in England, with one- or two-dose primary vaccination. METHODS: Integrated transmission-dynamic health-economic modeling explored the effects of targeting strategy, first- and second-dose uptake levels, and duration of vaccine protection, using observational estimates of vaccine protection. RESULTS: Vaccination with one or two primary doses is always cost-saving, irrespective of uptake, although vaccine sentiment is an important determinant of impact and cost-effectiveness. The most impactful and cost-effective targeting is offering "Vaccination-according-to-Risk" (VaR), to all patients with gonorrhea plus those reporting high numbers of sexual partners. If VaR is not feasible to implement then the more-restrictive strategy of "Vaccination-on-Diagnosis" (VoD) with gonorrhea is cost-effective, but much less impactful. Under conservative assumptions, VaR(2-dose) saves £7.62M(95%CrI:1.15-17.52) and gains 81.41(28.67-164.23) QALYs over 10 years; VoD(2-dose) saves £3.40M(0.48-7.71) and gains 41.26(17.52-78.25) QALYs versus no vaccination. Optimistic versus pessimistic vaccine-sentiment assumptions increase net benefits by ∼30%(VoD) or ∼60%(VaR). CONCLUSIONS: At UK costs, targeted 4CMenB vaccination of MSM gains QALYs and is cost-saving at any uptake level. Promoting uptake maximizes benefits and is an important role for behavioral science.

Mindful-Gratitude Practice Reduces Prejudice at High Levels of Collective Narcissism.

This research tested the hypothesis that mindful-gratitude practice attenuates the robust association between collective narcissism and prejudice. In Study 1 (a between-subjects study using a nationally representative sample of 569 Polish adults; 313 female), 10 min of mindful-gratitude practice-compared to mindful-attention practice and control-did not decrease prejudice (anti-Semitism), but weakened the positive link between collective narcissism and prejudice. In Study 2 (a preregistered, randomized, controlled-trial study using a convenience sample of 219 Polish adults; 168 female), a 6-week mobile app supported training in daily mindful-gratitude practice decreased prejudice (anti-Semitism, sexism, homophobia, anti-immigrant sentiment) and its link with collective narcissism compared to a wait-list control. The hypothesis-consistent results emphasize the social relevance of mindful-gratitude practice, a time- and cost-effective intervention.

Hybrid deep learning approach for sentiment analysis using text and emojis.

Sentiment Analysis (SA) is a technique for categorizing texts based on the sentimental polarity of people's opinions. This paper introduces a sentiment analysis (SA) model with text and emojis. The two preprocessed data's are data with text and emojis and text without emojis. Feature extraction consists text features and text with emojis features. The text features are features like N-grams, modified Term Frequency-Inverse Document Frequency (TF-IDF), and Bag-of-Words (BoW) features extracted from the text. In classification, CNN (Conventional Neural Network) and MLP (Multi-Layer Perception) use emojis and text-based SA. The CNN weight is optimized by a new Electric fish Customized Shark Smell Optimization (ECSSO) Algorithm. Similarly, the text-based SA is carried out by hybrid Long Short-Term Memory (LSTM) and Recurrent Neural Network (RNN) classifiers. The bagged data are given as input to the classification process via RNN and LSTM. Here, the weight of LSTM is optimized by the suggested ECSSO algorithm. Then, the mean of LSTM and RNN determines the final output. The specificity of the developed scheme is 29.01%, 42.75%, 23.88%,22.07%, 25.31%, 18.42%, 5.68%, 10.34%, 6.20%, 6.64%, and 6.84% better for 70% than other models. The efficiency of the proposed scheme is computed and evaluated.

Sentiment analysis using graph-based Quickprop method for product quality enhancement.

The degree to which customers express satisfaction with a product on Twitter and other social media platforms is increasingly used to evaluate product quality. However, the volume and variety of textual data make traditional sentiment analysis methods challenging. The nuanced and context-dependent nature of product-related opinions presents a challenge for existing tools. This research addresses this gap by utilizing complex graph-based modelling strategies to capture the intricacies of real-world data. The Graph-based Quickprop Method constructs a graph model using the Sentiment140 dataset with 1.6 million tweets, where individuals are nodes and interactions are edges. Experimental results show a significant increase in sentiment classification accuracy, demonstrating the method's efficacy. This contribution underscores the importance of relational structures in sentiment analysis and provides a robust framework for extracting actionable insights from user-generated content, leading to improved product quality evaluations. The GQP-PQE method advances sentiment analysis and offers practical implications for businesses seeking to enhance product quality through a better understanding of consumer feedback on social media.

Tree hierarchical deep convolutional neural network optimized with sheep flock optimization algorithm for sentiment classification of Twitter data.

The increasing volume of online reviews and tweets poses significant challenges for sentiment classification because of the difficulty in obtaining annotated training data. This paper aims to enhance sentiment classification of Twitter data by developing a robust model that improves classification accuracy and computational efficiency. The proposed method named Tree Hierarchical Deep Convolutional Neural Network optimized with Sheep Flock Optimization Algorithm for Sentiment Classification of Twitter Data (SCTD-THDCNN-SFOA) utilizes the Stanford Sentiment Treebank dataset. The process begins with pre-processing steps including Tokenization, Stop words Elimination, Filtering, Hashtag Removal, and Multiword Grouping. The Gray Level Co-occurrence Matrix Window Adaptive Algorithm is employed to extract features, such as emoticon counts, punctuation counts, gazetteer word existence, n-grams, and part of speech tags. These features are selected using Entropy-Kurtosis-based Feature Selection approach. Finally, the Tree Hierarchical Deep Convolutional Neural Network enhanced by the Sheep Flock Optimization Algorithm is used to categorize the Twitter data as positive, negative, and neutral sentiments. The proposed SCTD-THDCNN-SFOA method demonstrates superior performance, achieving higher accuracy and lesser computation time than the existing models, respectively. The SCTD-THDCNN-SFOA framework significantly improves the accuracy and efficiency of sentiment classification for Twitter data.

Public Perceptions of the Australian Health System During COVID-19: Findings From a 2021 Survey Compared to Four Previous Surveys.

BACKGROUND: This study examines the perceptions of the Australian public canvassed in 2021 during the COVID-19 pandemic about their health system compared to four previous surveys (2008, 2010, 2012 and 2018). METHODS: In 2021, a nationwide online survey was conducted with a representative sample of Australians (N = 5100) recruited via market research panels. The results were compared to previous nationwide Australian survey samples from 2018 (N = 1024), 2012 (N = 1200), 2010 (N = 1201) and 2008 (N = 1146). The survey included questions consistent with previous polls regarding self-reported health status and overall opinions of, and confidence in, the Australian health system. RESULTS: There was an increase in the proportion of respondents reporting positive perceptions at each survey between 2008 and 2021, with a significantly higher proportion of respondents expressing a more positive view of the Australian healthcare system in 2021 compared to previous years (χ2(8, N = 9645) = 487.63, p < 0.001). In 2021, over two-thirds of respondents (n = 3949/5100, 77.4%) reported that following the COVID-19 pandemic, their confidence in the Australian healthcare system had either remained the same (n = 2433/5100, 47.7%) or increased (n = 1516/5100, 29.7%). Overall, respondents living in regional or remote regions, younger Australians (< 45 years) and women held less positive views in relation to the system. In 2021, the most frequently identified area for urgent improvement was the need for more healthcare workers (n = 1350/3576, 37.8%), an area of concern particularly for Australians residing in regional or remote areas (n = 590/1385, 42.6%). CONCLUSIONS: Irrespective of disruptions to the Australian healthcare system caused by the COVID-19 pandemic, Australians' perceptions of their healthcare system were positive in 2021. However, concerns were raised about inadequate workforce capacity and the cost of healthcare, with differences identified by age groups and geographical location. PATIENT OR PUBLIC CONTRIBUTION: Health consumer representatives from the Consumers Health Forum of Australia contributed to the co-design, deployment, analysis and interpretation of the results of this survey. J.A. and L.W. from the Consumers Health Forum of Australia contributed to the development of the paper.

Mapping loneliness through comparative analysis of USA and India using social intelligence analysis.

Loneliness, a widespread global public health concern, has far-reaching implications for mental and physical well-being, as well as economic productivity. It also increases the risk of life-threatening conditions. This study conducts a comparative analysis of loneliness in the USA and India using Twitter data, aiming to contribute to a global public health map on loneliness. Collecting 4.1 million tweets globally in October 2022 containing keywords like "lonely", "loneliness", and "alone", the analysis focuses on sentiment and psychosocial linguistic features. Utilizing the Valence Aware Dictionary for Sentiment Reasoning (VADER) for sentiment analysis, the study explores variations in loneliness dynamics across cities, revealing geographical distinctions in correlated topics. The tweets with negative sentiment were further analyzed for psychosocial linguistic features to find a meaningful correlation between loneliness and socioeconomic and emotional themes and factors. Results give detailed top correlated topics with loneliness for each city. The results showed that the dynamics of loneliness through the topics correlated vary across geographical locations. Social media data can be used to capture the dynamics of loneliness which can vary from one place to another depending on the socioeconomic and cultural norms and sociopolitical policies. Social media data to understand loneliness can also provide useful information and insight for public health and policymaking.

Contents analysis of thyroid cancer-related information uploaded to YouTube by physicians in Korea: endorsing thyroid cancer screening, potentially leading to overdiagnosis.

BACKGROUND: Thyroid cancer overdiagnosis is a major public health issue in South Korea, which has the highest incidence rate. The accessibility of information through the Internet, particularly on YouTube, could potentially impact excessive screening. This study aimed to analyze the content of thyroid cancer-related YouTube videos, particularly those from 2016 onwards, to evaluate the potential spread of misinformation. METHODS: A total of 326 videos for analysis were collected using a video search protocol with the keyword "thyroid cancer" on YouTube. This study classified the selected YouTube videos as either provided by medical professionals or not and used topic clustering with LDA (latent dirichlet allocation), sentiment analysis with KoBERT (Korean bidirectional encoder representations from transformers), and reliability evaluation to analyze the content. The proportion of mentions of poor prognosis for thyroid cancer and the categorization of advertising content was also analyzed. RESULTS: Videos by medical professionals were categorized into 7 topics, with "Thyroid cancer is not a 'Good cancer'" being the most common. The number of videos opposing excessive thyroid cancer screening decreased gradually yearly. Videos advocating screening received more favorable comments from viewers than videos opposing excessive thyroid cancer screening. Patient experience videos were categorized into 6 topics, with the "Treatment process and after-treatment" being the most common. CONCLUSION: This study found that a significant proportion of videos uploaded by medical professionals on thyroid cancer endorse the practice, potentially leading to excessive treatments. The study highlights the need for medical professionals to provide high-quality and unbiased information on social media platforms to prevent the spread of medical misinformation and the need for criteria to judge the content and quality of online health information.

Unveiling public perceptions at the beginning of lockdown: an application of structural topic modeling and sentiment analysis in the UK and India.

BACKGROUND: The appearance of the COVID-19 virus in December 2019, quickly escalated into a global crisis, prompting the World Health Organization to recommend regional lockdowns. While effective in curbing the virus's spread, these measures have triggered intense debates on social media platforms, exposing widespread public anxiety and skepticism. The spread of fake news further fueled public unrest and negative emotions, potentially undermining the effectiveness of anti-COVID-19 policies. Exploring the narratives surrounding COVID-19 on social media immediately following the lockdown announcements presents an intriguing research avenue. The purpose of this study is to examine social media discourse to identify the topics discussed and, more importantly, to analyze differences in the focus and emotions expressed by the public in two countries (the UK and India). This is done with an analysis of a big corpus of tweets. METHODS: The datasets comprised of COVID-19-related tweets in English, published between March 29th and April 11th 2020 from residents in the UK and India. Methods employed in the analysis include identification of latent topics and themes, assessment of the popularity of tweets on topic distributions, examination of the overall sentiment, and investigation of sentiment in specific topics and themes. RESULTS: Safety measures, government responses and cooperative supports are common themes in the UK and India. Personal experiences and cooperations are top discussion for both countries. The impact on specific groups is given the least emphasis in the UK, whereas India places the least focus on discussions related to social media and news reports. Supports, discussion about the UK PM Boris Johnson and appreciation are strong topics among British popular tweets, whereas confirmed cases are discussed most among Indian popular tweets. Unpopular tweets in both countries pay the most attention to issues regarding lockdown. According to overall sentiment, positive attitudes are dominated in the UK whilst the sentiment is more neutral in India. Trust and anticipation are the most prevalent emotions in both countries. In particular, the British population felt positive about community support and volunteering, personal experiences, and government responses, while Indian people felt positive about cooperation, government responses, and coping strategies. Public health situations raise negative sentiment both in the UK and India. CONCLUSIONS: The study emphasizes the role of cultural values in crisis communication and public health policy. Individualistic societies prioritize personal freedom, requiring a balance between individual liberty and public health measures. Collectivistic societies focus on community impact, suggesting policies that could utilize community networks for public health compliance. Social media shapes public discourse during pandemics, with popular and unpopular tweets reflecting and reshaping discussions. The presence of fake news may distort topics of high public interest, necessitating authenticity confirmation by official bloggers. Understanding public concerns and popular content on social media can help authorities tailor crisis communication to improve public engagement and health measure compliance.

The rise and fall of rationality in language.

The surge of post-truth political argumentation suggests that we are living in a special historical period when it comes to the balance between emotion and reasoning. To explore if this is indeed the case, we analyze language in millions of books covering the period from 1850 to 2019 represented in Google nGram data. We show that the use of words associated with rationality, such as "determine" and "conclusion," rose systematically after 1850, while words related to human experience such as "feel" and "believe" declined. This pattern reversed over the past decades, paralleled by a shift from a collectivistic to an individualistic focus as reflected, among other things, by the ratio of singular to plural pronouns such as "I"/"we" and "he"/"they." Interpreting this synchronous sea change in book language remains challenging. However, as we show, the nature of this reversal occurs in fiction as well as nonfiction. Moreover, the pattern of change in the ratio between sentiment and rationality flag words since 1850 also occurs in New York Times articles, suggesting that it is not an artifact of the book corpora we analyzed. Finally, we show that word trends in books parallel trends in corresponding Google search terms, supporting the idea that changes in book language do in part reflect changes in interest. All in all, our results suggest that over the past decades, there has been a marked shift in public interest from the collective to the individual, and from rationality toward emotion.

Differences in Fear and Negativity Levels Between Formal and Informal Health-Related Websites: Analysis of Sentiments and Emotions.

BACKGROUND: Searching for web-based health-related information is frequently performed by the public and may affect public behavior regarding health decision-making. Particularly, it may result in anxiety, erroneous, and harmful self-diagnosis. Most searched health-related topics are cancer, cardiovascular diseases, and infectious diseases. A health-related web-based search may result in either formal or informal medical website, both of which may evoke feelings of fear and negativity. OBJECTIVE: Our study aimed to assess whether there is a difference in fear and negativity levels between information appearing on formal and informal health-related websites. METHODS: A web search was performed to retrieve the contents of websites containing symptoms of selected diseases, using selected common symptoms. Retrieved websites were classified into formal and informal websites. Fear and negativity of each content were evaluated using 3 transformer models. A fourth transformer model was fine-tuned using an existing emotion data set obtained from a web-based health community. For formal and informal websites, fear and negativity levels were aggregated. t tests were conducted to evaluate the differences in fear and negativity levels between formal and informal websites. RESULTS: In this study, unique websites (N=1448) were collected, of which 534 were considered formal and 914 were considered informal. There were 1820 result pages from formal websites and 1494 result pages from informal websites. According to our findings, fear levels were statistically higher (t2753=3.331; P<.001) on formal websites (mean 0.388, SD 0.177) than on informal websites (mean 0.366, SD 0.168). The results also show that the level of negativity was statistically higher (t2753=2.726; P=.006) on formal websites (mean 0.657, SD 0.211) than on informal websites (mean 0.636, SD 0.201). CONCLUSIONS: Positive texts may increase the credibility of formal health websites and increase their usage by the general public and the public's compliance to the recommendations. Increasing the usage of natural language processing tools before publishing health-related information to achieve a more positive and less stressful text to be disseminated to the public is recommended.

Investigation of Public Acceptance of Misinformation Correction in Social Media Based on Sentiment Attributions: Infodemiology Study Using Aspect-Based Sentiment Analysis.

BACKGROUND: The proliferation of misinformation on social media is a significant concern due to its frequent occurrence and subsequent adverse social consequences. Effective interventions for and corrections of misinformation have become a focal point of scholarly inquiry. However, exploration of the underlying causes that affect the public acceptance of misinformation correction is still important and not yet sufficient. OBJECTIVE: This study aims to identify the critical attributions that influence public acceptance of misinformation correction by using attribution analysis of aspects of public sentiment, as well as investigate the differences and similarities in public sentiment attributions in different types of misinformation correction. METHODS: A theoretical framework was developed for analysis based on attribution theory, and public sentiment attributions were divided into 6 aspects and 11 dimensions. The correction posts for the 31 screened misinformation events comprised 33,422 Weibo posts, and the corresponding Weibo comments amounted to 370,218. A pretraining model was used to assess public acceptance of misinformation correction from these comments, and the aspect-based sentiment analysis method was used to identify the attributions of public sentiment response. Ultimately, this study revealed the causality between public sentiment attributions and public acceptance of misinformation correction through logistic regression analysis. RESULTS: The findings were as follows: First, public sentiments attributed to external attribution had a greater impact on public acceptance than those attributed to internal attribution. The public associated different aspects with correction depending on the type of misinformation. The accuracy of the correction and the entity responsible for carrying it out had a significant impact on public acceptance of misinformation correction. Second, negative sentiments toward the media significantly increased, and public trust in the media significantly decreased. The collapse of media credibility had a detrimental effect on the actual effectiveness of misinformation correction. Third, there was a significant difference in public attitudes toward the official government and local governments. Public negative sentiments toward local governments were more pronounced. CONCLUSIONS: Our findings imply that public acceptance of misinformation correction requires flexible communication tailored to public sentiment attribution. The media need to rebuild their image and regain public trust. Moreover, the government plays a central role in public acceptance of misinformation correction. Some local governments need to repair trust with the public. Overall, this study offered insights into practical experience and a theoretical foundation for controlling various types of misinformation based on attribution analysis of public sentiment.

Efficacy of ChatGPT in Cantonese Sentiment Analysis: Comparative Study.

BACKGROUND: Sentiment analysis is a significant yet difficult task in natural language processing. The linguistic peculiarities of Cantonese, including its high similarity with Standard Chinese, its grammatical and lexical uniqueness, and its colloquialism and multilingualism, make it different from other languages and pose additional challenges to sentiment analysis. Recent advances in models such as ChatGPT offer potential viable solutions. OBJECTIVE: This study investigated the efficacy of GPT-3.5 and GPT-4 in Cantonese sentiment analysis in the context of web-based counseling and compared their performance with other mainstream methods, including lexicon-based methods and machine learning approaches. METHODS: We analyzed transcripts from a web-based, text-based counseling service in Hong Kong, including a total of 131 individual counseling sessions and 6169 messages between counselors and help-seekers. First, a codebook was developed for human annotation. A simple prompt ("Is the sentiment of this Cantonese text positive, neutral, or negative? Respond with the sentiment label only.") was then given to GPT-3.5 and GPT-4 to label each message's sentiment. GPT-3.5 and GPT-4's performance was compared with a lexicon-based method and 3 state-of-the-art models, including linear regression, support vector machines, and long short-term memory neural networks. RESULTS: Our findings revealed ChatGPT's remarkable accuracy in sentiment classification, with GPT-3.5 and GPT-4, respectively, achieving 92.1% (5682/6169) and 95.3% (5880/6169) accuracy in identifying positive, neutral, and negative sentiment, thereby outperforming the traditional lexicon-based method, which had an accuracy of 37.2% (2295/6169), and the 3 machine learning models, which had accuracies ranging from 66% (4072/6169) to 70.9% (4374/6169). CONCLUSIONS: Among many text analysis techniques, ChatGPT demonstrates superior accuracy and emerges as a promising tool for Cantonese sentiment analysis. This study also highlights ChatGPT's applicability in real-world scenarios, such as monitoring the quality of text-based counseling services and detecting message-level sentiments in vivo. The insights derived from this study pave the way for further exploration into the capabilities of ChatGPT in the context of underresourced languages and specialized domains like psychotherapy and natural language processing.

From Tweets to Streets: Observational Study on the Association Between Twitter Sentiment and Anti-Asian Hate Crimes in New York City from 2019 to 2022.

BACKGROUND: Anti-Asian hate crimes escalated during the COVID-19 pandemic; however, limited research has explored the association between social media sentiment and hate crimes toward Asian communities. OBJECTIVE: This study aims to investigate the relationship between Twitter (rebranded as X) sentiment data and the occurrence of anti-Asian hate crimes in New York City from 2019 to 2022, a period encompassing both before and during COVID-19 pandemic conditions. METHODS: We used a hate crime dataset from the New York City Police Department. This dataset included detailed information on the occurrence of anti-Asian hate crimes at the police precinct level from 2019 to 2022. We used Twitter's application programming interface for Academic Research to collect a random 1% sample of publicly available Twitter data in New York State, including New York City, that included 1 or more of the selected Asian-related keywords and applied support vector machine to classify sentiment. We measured sentiment toward the Asian community using the rates of negative and positive sentiment expressed in tweets at the monthly level (N=48). We used negative binomial models to explore the associations between sentiment levels and the number of anti-Asian hate crimes in the same month. We further adjusted our models for confounders such as the unemployment rate and the emergence of the COVID-19 pandemic. As sensitivity analyses, we used distributed lag models to capture 1- to 2-month lag times. RESULTS: A point increase of 1% in negative sentiment rate toward the Asian community in the same month was associated with a 24% increase (incidence rate ratio [IRR] 1.24; 95% CI 1.07-1.44; P=.005) in the number of anti-Asian hate crimes. The association was slightly attenuated after adjusting for unemployment and COVID-19 emergence (ie, after March 2020; P=.008). The positive sentiment toward Asian tweets with a 0-month lag was associated with a 12% decrease (IRR 0.88; 95% CI 0.79-0.97; P=.002) in expected anti-Asian hate crimes in the same month, but the relationship was no longer significant after adjusting for the unemployment rate and the emergence of COVID-19 pandemic (P=.11). CONCLUSIONS: A higher negative sentiment level was associated with more hate crimes specifically targeting the Asian community in the same month. The findings highlight the importance of monitoring public sentiment to predict and potentially mitigate hate crimes against Asian individuals.

Uncovering the Complexity of Perinatal Polysubstance Use Disclosure Patterns on X: Mixed Methods Study.

BACKGROUND: According to the Morbidity and Mortality Weekly Report, polysubstance use among pregnant women is prevalent, with 38.2% of those who consume alcohol also engaging in the use of one or more additional substances. However, the underlying mechanisms, contexts, and experiences of polysubstance use are unclear. Organic information is abundant on social media such as X (formerly Twitter). Traditional quantitative and qualitative methods, as well as natural language processing techniques, can be jointly used to derive insights into public opinions, sentiments, and clinical and public health policy implications. OBJECTIVE: Based on perinatal polysubstance use (PPU) data that we extracted on X from May 1, 2019, to October 31, 2021, we proposed two primary research questions: (1) What is the overall trend and sentiment of PPU discussions on X? (2) Are there any distinct patterns in the discussion trends of PPU-related tweets? If so, what are the implications for perinatal care and associated public health policies? METHODS: We used X's application programming interface to extract >6 million raw tweets worldwide containing ≥2 prenatal health- and substance-related keywords provided by our clinical team. After removing all non-English-language tweets, non-US tweets, and US tweets without disclosed geolocations, we obtained 4848 PPU-related US tweets. We then evaluated them using a mixed methods approach. The quantitative analysis applied frequency, trend analysis, and several natural language processing techniques such as sentiment analysis to derive statistics to preview the corpus. To further understand semantics and clinical insights among these tweets, we conducted an in-depth thematic content analysis with a random sample of 500 PPU-related tweets with a satisfying κ score of 0.7748 for intercoder reliability. RESULTS: Our quantitative analysis indicates the overall trends, bigram and trigram patterns, and negative sentiments were more dominant in PPU tweets (2490/4848, 51.36%) than in the non-PPU sample (1323/4848, 27.29%). Paired polysubstance use (4134/4848, 85.27%) was the most common, with the combination alcohol and drugs identified as the most mentioned. From the qualitative analysis, we identified 3 main themes: nonsubstance, single substance, and polysubstance, and 4 subthemes to contextualize the rationale of underlying PPU behaviors: lifestyle, perceptions of others' drug use, legal implications, and public health. CONCLUSIONS: This study identified underexplored, emerging, and important topics related to perinatal PPU, with significant stigmas and legal ramifications discussed on X. Overall, public sentiments on PPU were mixed, encompassing negative (2490/4848, 51.36%), positive (1884/4848, 38.86%), and neutral (474/4848, 9.78%) sentiments. The leading substances in PPU were alcohol and drugs, and the normalization of PPU discussed on X is becoming more prevalent. Thus, this study provides valuable insights to further understand the complexity of PPU and its implications for public health practitioners and policy makers to provide proper access and support to individuals with PPU.

Public Opinion About COVID-19 on a Microblog Platform in China: Topic Modeling and Multidimensional Sentiment Analysis of Social Media.

BACKGROUND: The COVID-19 pandemic raised wide concern from all walks of life globally. Social media platforms became an important channel for information dissemination and an effective medium for public sentiment transmission during the COVID-19 pandemic. OBJECTIVE: Mining and analyzing social media text information can not only reflect the changes in public sentiment characteristics during the COVID-19 pandemic but also help the government understand the trends in public opinion and reasonably control public opinion. METHODS: First, this study collected microblog comments related to the COVID-19 pandemic as a data set. Second, sentiment analysis was carried out based on the topic modeling method combining latent Dirichlet allocation (LDA) and Bidirectional Encoder Representations from Transformers (BERT). Finally, a machine learning linear regression (ML-LR) model combined with a sparse matrix was proposed to explore the evolutionary trend in public opinion on social media and verify the high accuracy of the model. RESULTS: The experimental results show that, in different stages, the characteristics of public emotion are different, and the overall trend is from negative to positive. CONCLUSIONS: The proposed method can effectively reflect the characteristics of the different times and space of public opinion. The results provide theoretical support and practical reference in response to public health and safety events.

Using Natural Language Processing to Explore Social Media Opinions on Food Security: Sentiment Analysis and Topic Modeling Study.

BACKGROUND: Social media has the potential to be of great value in understanding patterns in public health using large-scale analysis approaches (eg, data science and natural language processing [NLP]), 2 of which have been used in public health: sentiment analysis and topic modeling; however, their use in the area of food security and public health nutrition is limited. OBJECTIVE: This study aims to explore the potential use of NLP tools to gather insights from real-world social media data on the public health issue of food security. METHODS: A search strategy for obtaining tweets was developed using food security terms. Tweets were collected using the Twitter application programming interface from January 1, 2019, to December 31, 2021, filtered for Australia-based users only. Sentiment analysis of the tweets was performed using the Valence Aware Dictionary and Sentiment Reasoner. Topic modeling exploring the content of tweets was conducted using latent Dirichlet allocation with BigML (BigML, Inc). Sentiment, topic, and engagement (the sum of likes, retweets, quotations, and replies) were compared across years. RESULTS: In total, 38,070 tweets were collected from 14,880 Twitter users. Overall, the sentiment when discussing food security was positive, although this varied across the 3 years. Positive sentiment remained higher during the COVID-19 lockdown periods in Australia. The topic model contained 10 topics (in order from highest to lowest probability in the data set): "Global production," "Food insecurity and health," "Use of food banks," "Giving to food banks," "Family poverty," "Food relief provision," "Global food insecurity," "Climate change," "Australian food insecurity," and "Human rights." The topic "Giving to food banks," which focused on support and donation, had the highest proportion of positive sentiment, and "Global food insecurity," which covered food insecurity prevalence worldwide, had the highest proportion of negative sentiment. When compared with news, there were some events, such as COVID-19 support payment introduction and bushfires across Australia, that were associated with high periods of positive or negative sentiment. Topics related to food insecurity prevalence, poverty, and food relief in Australia were not consistently more prominent during the COVID-19 pandemic than before the pandemic. Negative tweets received substantially higher engagement across 2019 and 2020. There was no clear relationship between topics that were more likely to be positive or negative and have higher or lower engagement, indicating that the identified topics are discrete issues. CONCLUSIONS: In this study, we demonstrated the potential use of sentiment analysis and topic modeling to explore evolution in conversations on food security using social media data. Future use of NLP in food security requires the context of and interpretation by public health experts and the use of broader data sets, with the potential to track dimensions or events related to food security to inform evidence-based decision-making in this area.