Article
Computer Science, Artificial Intelligence
Mohammad Majid Abedi, Emanuele Sacchi
Summary: This research aims to extract, classify, and study drivers' affective states from road safety-related tweets using keyword filtering, geo-boundaries, natural language processing, and machine-learning classification. The results showed that the trained RF model with count vector, and SVM classifier with word-level TF-IDF performed best in separating road safety-related from unrelated tweets and determining the proposed classification tags.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Business
Vishanth Weerakkody, Uthayasankar Sivarajah, Kamran Mahroof, Takao Maruyama, Shan Lu
Summary: Business leaders and policymakers in service economies are increasingly focusing on promoting well-being among workers. Well-being can drive economic growth, but also have adverse effects if neglected. Therefore, enhancing subjective well-being is crucial for sustainable development.
JOURNAL OF BUSINESS RESEARCH
(2021)
Article
Green & Sustainable Science & Technology
Blanka Klimova, Marcel Pikhart, Szymon Dziuba, Anna Cierniak-Emerych
Summary: Learning a foreign language may contribute to the subjective well-being of older individuals, maintaining it at a relatively high level. Factor analysis revealed the correlation between foreign language learning and seniors' subjective well-being.
Article
Green & Sustainable Science & Technology
Na Ke, Guoqing Shi, Ying Zhou
Summary: This study uses machine learning techniques to propose a stacking model to predict the subjective well-being of Chinese residents and reveal changes in important factors affecting SWB. The results show that the stacking model outperforms traditional models and contributes to a more harmonious society.
Article
Agricultural Engineering
Chao-Tung Yang, Endah Kristiani, Yoong Kit Leong, Jo-Shu Chang
Summary: This paper examines and summarizes the literature related to artificial intelligence (AI) in the bioprocessing field, aiming to explore the potential of machine learning algorithms in revolutionizing bioengineering. By employing natural language processing (NLP), papers from 2013 to 2022 with specific keywords of bioprocessing using AI were collected and analyzed. The results show that in the past five years, 50% of the publications focused on topics such as hybrid models, artificial neural networks (ANN), biopharmaceutical manufacturing, and biorefinery. The summarization and analysis indicate that implementing AI can improve the design and process engineering strategies in bioprocessing fields.
BIORESOURCE TECHNOLOGY
(2023)
Article
Economics
Ekaterina Oparina, Sorawoot Srisuma
Summary: The study used novel nonparametric techniques to identify measurement error in reported life satisfaction (LS) and found that the bias is not severe enough to distort the main drivers of LS. However, there is an important difference where women tend to over-report their latent LS relative to men, providing insight into the gender puzzle of reported happiness levels.
JOURNAL OF BUSINESS & ECONOMIC STATISTICS
(2022)
Article
Computer Science, Artificial Intelligence
Sergey Smetanin, Mikhail Komarov
Summary: This study utilized historical posts from social networks to measure subjective well-being, and conducted sentiment analysis on the Russian segment of Twitter. The calculated well-being indicators showed moderate correlation with traditional survey-based affect indices in Russia.
PEERJ COMPUTER SCIENCE
(2022)
Article
Mathematics
Sergey Smetanin
Summary: This study proposed a formal model for calculating the observable subjective well-being (OSWB) indicator based on social network posts, achieving state-of-the-art results in sentiment analysis and identifying daily and weekly patterns in happiness levels among the Russian population. Different demographic groups showed varying levels of happiness on a daily, weekly, and monthly basis, confirming the significance of post-stratification in OSWB studies.
Article
Food Science & Technology
Haoyang Zhang, Dachuan Zhang, Zhisheng Wei, Yan Li, Shaji Wu, Zhiheng Mao, Chunmeng He, Haorui Ma, Xin Zeng, Xiaoling Xie, Xingran Kou, Bingwen Zhang
Summary: In order to systematically collect and analyze public opinion on food safety in Greater China, the researchers developed IFoodCloud, which automatically gathers data from over 3,100 public sources. By constructing sentiment classification models using various lexicon-based and machine learning-based algorithms integrated with IFoodCloud, they were able to rapidly understand public sentiment towards specific food safety incidents. Their best model achieved an impressive F1 score of 0.9737, indicating its strong predictive ability and reliability. Through the analysis of public sentiment, they demonstrated the potential of using big data and machine learning for risk communication and decision-making during the early stages of the 2019 Coronavirus Disease pandemic.
CURRENT RESEARCH IN FOOD SCIENCE
(2023)
Article
Ecology
Yian Lin, Spencer A. Wood, Joshua J. Lawler
Summary: This study used geolocated tweets from Seattle to investigate the relationship between subjective well-being and natural environments. The results showed that the relationship varied depending on the location in the built environment, highlighting the complexity of this relationship.
LANDSCAPE AND URBAN PLANNING
(2022)
Article
Green & Sustainable Science & Technology
Lu Cheng, Zhifu Mi, Yi-Ming Wei, Shidong Wang, Klaus Hubacek
Summary: The self-reported life satisfaction of China's population has not improved as expected during the economic boom, possibly due to environmental pollution. This study finds a negative correlation between air pollution and subjective well-being, with a more significant decline in well-being during hot seasons. Additionally, residents in wealthier regions are more sensitive to air pollution.
JOURNAL OF CLEANER PRODUCTION
(2022)
Article
Chemistry, Multidisciplinary
Murat Dener, Gokce Ok, Abdullah Orman
Summary: The study suggests using memory data in malware detection and applying deep learning and machine learning approaches in a big data environment. Results show that the Logistic Regression algorithm achieved the most successful malware detection in memory analysis.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Krzysztof Fiok, Waldemar Karwowski, Edgar Gutierrez, Maciej Wilamowski
Summary: The research in sentiment analysis has been thriving with the advancement of machine learning and deep learning. By comparing with SemEval, the study evaluates performance of popular natural language processing methods and explores how new unsupervised ML techniques can enhance predictive performance of models.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Public, Environmental & Occupational Health
Xiaoyan Huang, Chenchen Kang, Chun Yin, Yu Li
Summary: This study investigates the relative importance of urban and individual attributes to subjective well-being (SWB) and their nonlinear associations using a machine learning approach. The results show that income and age are the most important predictors, and urban facilities have a larger contribution than urban development factors. Additionally, urban attributes have nonlinear and threshold effects on SWB, with cultural facilities and green space having an inverted U-shaped correlation, while educational facilities, medical facilities, and population size having monotonic associations with SWB and specific thresholds.
FRONTIERS IN PUBLIC HEALTH
(2023)
Article
Health Care Sciences & Services
Shuijin Li, Tingshao Zhu
Summary: This study uses big data modelling with microblogs to examine the relationship between China's Livelihood Index and subjective well-being. The analysis provides direct validation of the influence of the Livelihood Index on subjective well-being. The research finds that there is Granger causality running from the Livelihood Index to subjective well-being, indicating a positive correlation between the two. Additionally, the study shows Granger causality between a life stress indicator and a life satisfaction indicator. The findings also indicate positive correlations between the education indicator model and life satisfaction, as well as positive relationships between medical and health indicators and life satisfaction. Conversely, there is a negative correlation between the traffic indicator model and life satisfaction. The study further reveals bidirectional Granger causality and a positive correlation between economic development and subjective well-being. However, the correlation between the Livelihood Index and economic development appears weaker in regions with stronger economic growth. The study suggests measures to improve subjective well-being, including increasing gross domestic product per capita and absolute per capita income, as well as enhancing medical and health services, alleviating traffic congestion, improving the teacher-student ratio, and increasing education universalisation rate. These steps would promote equitable and balanced development of the Livelihood Index across China's provinces.
Article
Psychology, Clinical
Youngseo Son, Sean A. P. Clouston, Roman Kotov, Johannes C. Eichstaedt, Evelyn J. Bromet, Benjamin J. Luft, H. Andrew Schwartz
Summary: This study demonstrates the value of AI in understanding PTSD in a vulnerable population. Future studies should extend this application to other trauma exposures and to other demographic groups, especially under-represented minorities.
PSYCHOLOGICAL MEDICINE
(2023)
Article
Psychology, Multidisciplinary
David B. Yaden, Salvatore Giorgi, Margaret L. Kern, Alejandro Adler, Lyle H. Ungar, Martin E. P. Seligman, Johannes C. Eichstaedt
Summary: Religion and spirituality are multidimensional constructs involving practices, rituals, and experiences. This study explores the language associations of different dimensions and finds divergent profiles. Additionally, it reveals that non-believers focus more on emotions such as inspiration and gratitude rather than religious doctrine.
PSYCHOLOGY OF RELIGION AND SPIRITUALITY
(2023)
Article
Multidisciplinary Sciences
Kokil Jaidka
Summary: Spatial aggregates of survey and web search data reveal the heterogeneous well-being effects of social media platforms. The study finds that frequent visits to Facebook have consistently positive well-being effects, while visits to Instagram have negative effects. Furthermore, these effects vary across different population groups, with white and high-income individuals experiencing more positive effects and younger and Black populations experiencing adverse effects.
SCIENTIFIC REPORTS
(2022)
Article
Psychology, Social
Jeremy A. Frimer, Harinder Aujla, Matthew Feinberg, Linda J. Skitka, Karl Aquino, Johannes C. Eichstaedt, Robb Willer
Summary: This study provides the first systematic investigation of the incivility trends of American politicians on Twitter. The research reveals a 23% increase in incivility over the past decade on Twitter, partly driven by politicians engaging in greater incivility following positive feedback. Uncivil tweets tend to receive more approval and attention, leading to more uncivil tweets thereafter.
SOCIAL PSYCHOLOGICAL AND PERSONALITY SCIENCE
(2023)
Article
Communication
Subhayan Mukerjee, Kokil Jaidka, Yphtach Lelkes
Summary: Prior research suggests that Twitter users in the United States are more politically engaged and partisan than the general American population. However, this study finds little evidence to support this claim. The study analyzes the most popular Twitter accounts in the United States and concludes that the platform is not as political as previously thought. Ordinary Americans are more likely to follow nonpolitical opinion leaders on Twitter, and there is no significant polarization among these opinion leaders.
POLITICAL COMMUNICATION
(2022)
Letter
Psychology, Biological
Sean Fischer, Kokil Jaidka, Yphtach Lelkes
NATURE HUMAN BEHAVIOUR
(2022)
Review
Psychology, Developmental
Katharine Lancaster, Anoo Bhopti, Margaret L. Kern, Rachel Taylor, Annick Janson, Katherine Harding
Summary: This systematic review examines the quantitative evidence from the past decade on the effectiveness of peer support programmes in improving the well-being and quality of life for parents/carers of children with disability/chronic illnesses. The results suggest that peer support is effective in reducing distress and improving well-being and quality of life for parents, but the included studies have limitations in terms of bias.
CHILD CARE HEALTH AND DEVELOPMENT
(2023)
Article
Communication
Kokil Jaidka, Subhayan Mukerjee, Yphtach Lelkes
Summary: Algorithms play a critical role in directing online attention on social media, which has raised concerns about perpetuating bias. This study examined shadowbanning on Twitter, where users or their content are temporarily hidden. By testing a stratified random sample of American Twitter accounts, the study identified the factors predicting shadowbans. The findings showed that accounts with bot-like behavior were more likely to be shadowbanned, while verified accounts were less likely. Offensive and politically-focused tweets also faced potential downgrading. These findings have implications for algorithmic accountability and future audits of social media platforms.
JOURNAL OF COMMUNICATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Yifei Wang, Kokil Jaidka
Summary: In an increasingly digitalized world, online information-seeking behaviors play a critical role in synthesizing public opinion. This study investigates whether search strategies align with the expected confirmation biases of regions with different partisan beliefs. The results show significant differences in search keywords adopted by Democrat-majority and Republican-majority regions in the United States, and suggest that the preferential use of certain search keywords can predict climate opinions.
SOCIAL SCIENCE COMPUTER REVIEW
(2023)
Article
Public, Environmental & Occupational Health
Sophia Lou, Salvatore Giorgi, Tingting Liu, Johannes C. Eichstaedt, Brenda Curtis
Summary: Extensive evidence shows that area-based disadvantage has negative effects on various life outcomes, including increased mortality and low economic mobility. However, the measurement of disadvantage using composite indices is inconsistent across studies. To address this issue, we compared 5 U.S. disadvantage indices at the county-level and their relationships with 24 diverse life outcomes. The Area Deprivation Index (ADI) and Child Opportunity Index 2.0 (COI) were found to be most related to a wide range of life outcomes, particularly physical health.
Article
Multidisciplinary Sciences
Salvatore Giorgi, David B. Yaden, Johannes C. Eichstaedt, Lyle H. Ungar, H. Andrew Schwartz, Amy Kwarteng, Brenda Curtis
Summary: Opioid poisoning is a major public health crisis in the United States, responsible for 75% of the drug-related deaths. A lack of measurement tools for social and psychological factors hinder research in this area. This study uses a multi-modal data set, including Twitter language, psychometric self-reports, and area-based measures, to predict and understand opioid poisoning. The results show that Twitter language predicted opioid poisoning mortality better than socio-demographics, healthcare access, physical pain, and psychological well-being factors.
SCIENTIFIC REPORTS
(2023)
Article
Psychology, Experimental
David B. Yaden, Salvatore Giorgi, Matthew Jordan, Anneke Buffone, Johannes C. Eichstaedt, H. Andrew Schwartz, Lyle Ungar, Paul Bloom
Summary: Many scholars argue that empathy is crucial for other-regarding sentiments and plays a significant role in our moral lives, while compassion is also seen as a relevant force for prosocial motivation and action. This study, using computational linguistics, explores the relationship between empathy and compassion. Analysis of Facebook posts reveals that individuals high in empathy use different language than those high in compassion, and high empathy without compassion is associated with negative health outcomes, while high compassion without empathy is related to positive health outcomes, lifestyle choices, and charitable giving. These findings support a moral motivation grounded in compassion rather than empathy.
Review
Psychology, Multidisciplinary
S. Sametoglu, D. H. M. Pelt, J. C. Eichstaedt, L. H. Ungar, M. Bartels
Summary: This article presents a systematic review and meta-analysis of studies on the effectiveness of social media text mining in measuring well-being. The results show that there is a correlation between social media text mining and survey-based assessments of well-being, making it a valuable tool for evaluating individual and regional well-being. The article also provides recommendations for future research, including considering language diversity and careful selection of data collection methods.
JOURNAL OF POSITIVE PSYCHOLOGY
(2023)
Article
Psychology, Applied
Selim Sametoglu, Dirk H. M. Pelt, Johannes C. Eichstaedt, Lyle H. Ungar, Meike Bartels
Summary: Wellbeing can be measured through surveys and social media text mining (SMTM). Comparing survey data and social media language features, the networks derived from both methods showed similar structures, consisting of five wellbeing dimensions. This suggests that survey and SMTM methods can complement each other to understand differences in human wellbeing.
APPLIED PSYCHOLOGY-HEALTH AND WELL BEING
(2023)
Article
Psychology, Clinical
Elizabeth C. Stade, Lyle Ungar, Johannes C. Eichstaedt, Garrick Sherman, Ayelet Meron Ruscio
Summary: Depression is associated with increased use of first-person pronouns and negative emotion words. However, previous studies have not differentiated between depression and anxiety. This study interviewed individuals with varying levels of depression and anxiety and found that certain language features are specific to depression, while others are specific to anxiety.
JOURNAL OF PSYCHOPATHOLOGY AND CLINICAL SCIENCE
(2023)
