Dual Consistency-enhanced Semi-supervised Sentiment Analysis towards COVID-19 Tweets

In the context of COVID-19, numerous people present their opinions through social networks. It is thus highly desired to conduct sentiment analysis towards COVID-19 tweets to learn the public's attitudes, and facilitate the government to make proper guidelines for avoiding the social unrest. Although many efforts have studied the text-based sentiment classification from various domains (e.g., delivery and shopping reviews), it is hard to directly use these classifiers for the sentiment analysis towards COVID-19 tweets due to the domain gap. In fact, developing the sentiment classifier for COVID-19 tweets is mainly challenged by the limited annotated training dataset, as well as the diverse and informal expressions of user-generated posts. To address these challenges, we construct a large-scale COVID-19 dataset from Weibo and propose a dual COnsistency-enhanced semi-superVIseD network for Sentiment Anlaysis (COVID-SA). In particular, we first introduce a knowledge-based augmentation method to augment data and enhance the model's robustness. We then employ BERT as the text encoder backbone for both labeled data, unlabeled data, and augmented data. Moreover, we propose a dual consistency (i.e., label-oriented consistency and instance-oriented consistency) regularization to promote the model performance. Extensive experiments on our self-constructed dataset and three public datasets show the superiority of COVID-SA over state-of-the-art baselines on various applications. IEEE

Young People Under ‘Finfluencer': The Rise of Financial Influencers on Instagram: An Abstract

Under the shadow of the covid-19 pandemic, millions of young people have the resulting consequences (e.g., increased uncertainty regarding their careers and economic prospects: McGrindle, 2020;Yuesti et al., 2020) started to think more seriously about how to manage their financial resources. As a result, personal finance has become trendy to speak about. Rather than proactive actions coming through the formal financial services industry to address the increasing demand for accessible financial information that does not require an advisor or costs a fortune, we are witnessing the uprising of personal finance influencers. These financial influencers (a.k.a. Finfluencers) talk about how to manage various aspects of financial life and do so by promoting their advice on a broad scope of money related topics (ranging from dealing with student loans, buying a house, how to ask for a pay raise, budgeting tips, and tricks, filing for tax returns, stock investment opportunities, and portfolio management) in short, light-hearted video formats that are posted on Instagram, Tiktok, YouTube and other social media platforms. Although currently, these Finfluencers can be regarded as a niche influencer marketing group, it is expected to grow exponentially in the coming years – especially now that designated sections on social media platforms are taking off (e.g., #FinTok, #FinTwit, #Finmeme, #StockTok, #Findependence). In line with this trend, this research explores how consumers consume social media influencers when they offer utilitarian (rather than hedonic) oriented products, services, and advice. Using text-mining techniques, we examine the content of personal finance influencers on Instagram to investigate how consumers respond to the recommendation of more utilitarian-oriented products. In addition, gender and race are an integral part of the consumer's perception of financial expertise. This research looks at the moderating role of gender and race on influencers' financial expertise on consumer response. We found that influencers with financial accreditations generate a more positive affective response compared with financial influencers who mainly share information stemming from personal experience. In addition, we found that influencers' gender and race moderate the relationship between financial expertise and consumers' affective responses significantly. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Short-term load forecasting based on machine learning and epidemic association features

Accurate power load forecasting is an important guarantee for normal operation of a power system. There have been problems of large fluctuations in load demand and difficulty in modeling historical reference load during the COVID-19 outbreak. Thus this paper proposes a short-term load forecasting method based on machine learning, silent index and rolling anxiety index. First, Google mobility data and epidemic data are used to construct the silent index and rolling anxiety index to quantify the impact of the economic and epidemic developments on the power load. Then, the maximal information coefficient is used to analyze the strong correlation factors of power load during the epidemic and introduce epidemic load correlation characteristics. Finally, meteorological data, historical load and the constructed epidemic correlation features are combined as the input variables of the prediction model, and the prediction algorithm is analyzed by multiple machine learning models. The results show that the load forecasting model with the introduction of the epidemic correlation features can effectively improve the accuracy of load forecasting during the epidemic. © 2022 Power System Protection and Control Press. All rights reserved.

Short-term load forecasting based on machine learning and epidemic association features

Accurate power load forecasting is an important guarantee for normal operation of a power system. There have been problems of large fluctuations in load demand and difficulty in modeling historical reference load during the COVID-19 outbreak. Thus this paper proposes a short-term load forecasting method based on machine learning, silent index and rolling anxiety index. First, Google mobility data and epidemic data are used to construct the silent index and rolling anxiety index to quantify the impact of the economic and epidemic developments on the power load. Then, the maximal information coefficient is used to analyze the strong correlation factors of power load during the epidemic and introduce epidemic load correlation characteristics. Finally, meteorological data, historical load and the constructed epidemic correlation features are combined as the input variables of the prediction model, and the prediction algorithm is analyzed by multiple machine learning models. The results show that the load forecasting model with the introduction of the epidemic correlation features can effectively improve the accuracy of load forecasting during the epidemic. © 2022 Power System Protection and Control Press. All rights reserved.

Comparative analysis of different scene for face mask detection based on YOLOv4

The 2019-nCoV can be transmitted through respiratory droplets and other methods, which greatly endangers public health security. Wearing masks correctly has been proven to be one of the effective means to prevent virus infection, but limited by the complexity of practical application scenarios, the wearing of masks still relies heavily on manual supervision. Therefore, a fast and accurate face mask wearing detection method is urgently needed. In this paper, a mask detection algorithm based on improved YOLO-v4 is proposed as a solution to the problems of low accuracy, poor real-time performance, and poor robustness caused by complicated environments. In addition, a number of different training approaches, such as mosaic data augment, CIOU, label smoothing, cosine annealing, etc., are introduced. These techniques help to increase the training speed of the model as well as the accuracy of its detection. With a fast-training model, the model will be able to detect and compare the results of samples from different scenarios. The experiment will compare front and side faces, different colored masks, scenes of varying complexity and other perspectives in a systematic way. The experiment's result was able to reach 99.38 % accuracy after the model was trained using data from a variety of face masks being worn. Experiment results, both quantitative and qualitative, indicate that the method can be adapted to most scenarios and offers effective ideas for improvement. © 2023 SPIE.

A perspective of historically cultural studies on Bob Dylan's epic "Murder Most Foul"

Bob Dylan released his newest ballad epic "Murder Most Foul" in the COVID-19 Pandemic 2020. He portrays himself as a poetic singer who cares about the history and destiny of his country and people by narrating the cultural and historical event of the 50s and 60s in the 20th century in the United States in retrospect. This article reflexes the culture, history, politics and globalization process in the pandemic by introducing concepts of collective representations, analogical modes of thinking and totality from a perspective of cultural studies in anticipating a united social community of the United States and even the whole world at large to forge ahead with courage and hope in a historically cultural sense.

Spatial-temporal characteristics and pollutant assessment in the early stage of COVID-19 outbreak in China

To explore the early stage spatial-temporal characteristics and to assess the factors of atmospheric pollution that may affect the development of coronavirus disease 2019（COVID-19）outbreak in the Chinese Mainland in 2020，we collected the daily new cases of COVID-19 in the Municipalities and Provinces from the websites of National and Provincial Health Commission of China. The spatiotemporal characteristics of COVID-19 epidemic were studied using autocorrelation analysis and trend analysis. The Spearman's correlation coefficient for ranked data and generalized additive model were used for risk assessment of air pollutants affecting the COVID-19 epidemic of Hubei Province. Daily new cases of COVID-19 in the Chinese Mainland totaled 39 877 from January 20th to February 9th of 2020. The global Moran index values of these three weeks were 0.249，0.307 and 0.297（P<0.01），respectively. There was a significant clustering phenomenon. The high incidence regions included Hunan Province，Guangdong Province，Jiangxi Province，Zhejiang Province，Anhui Province and Jiangsu Province. The epidemic hot spots were basically distributed in the area from 108° 47'-123° 10' E to 25° 31'-35° 20' N. Daily new cases of COVID-19 in Hubei Province was positively correlated with daily average concentrations of PM10，NO2 and O3 pollutants（ρ =0.515，0.579 and 0.536，P<0.05）. The lag effects of air pollutions were existed. The relative risk（RR）values of PM2.5and PM10 reached to maximum with lag0，the RR value of NO2 reached to maximum with lag4，and the RR value of O3 reached to maximum with lag 0~1. We estimated that a 10 μg/m3 increase in day-before NO2 daily average concentration was associated with a 32.745% （95% Confidence Interval（CI）：11.586%-57.916%）excess risk（ER）of daily new cases of COVID-19. And NO2 had a significant impact on daily new cases of COVID-19. When NO2 was introduced to PM2.5and PM10 separately，for every 10 μg/m3 rise in NO2 daily average concentration，the ER of daily new cases of COVID-19 was 23.929%（95% CI：4.705%-46.682%）and 24.672%（95% CI：5.379%-47.496%），respectively. The study showed that the southeast was the main spread direction in the early stage of COVID-19 outbreak in the Chinese Mainland in 2020. Reducing the atmospheric concentration of nitrogen dioxide in epidemic hot spots has a positive effect on epidemic prevention and control. © 2022 Journal of Zhongshan University. All rights reserved.

Multidisciplinary Engagement of Diverse Students in Computer Science Education through Research Focused on Social Media COVID-19 Misinformation

The ongoing COVID-19 pandemic has disrupted vital elements of personal and public health, society, and education. Increasingly with the viral pandemic, misinformation on health and science issues has been disseminated online. We developed an undergraduate training program focused on producing and presenting research to combat the rampant spread of this misinformation. Online misinformation represents a complex, multidisciplinary problem. Consequently, recruitment of students to the program was not exclusive to those from Computer Science or Science, Technology, Engineering, and Math (STEM) educational backgrounds. Participants were actively recruited from fields such as Linguistics, Social and Political sciences. This data analytics outreach program aimed to train educationally and demographically diverse undergraduate students in computational techniques and presentation skills through guided research regarding the current burst of misinformation. Over ten weeks, participants were instructed in an online curriculum covering five milestones: Python programming, data processing, machine learning with natural language processing, visualization, and presentation. Subsequently, participants were engaged in Computer Science research analyzing a real-world data set gathered from Twitter™ 1 between January and June 2020. Participants were organized into teams to investigate subtopics within the broader subject of misinformation: 1) detecting social media bot accounts, 2) identifying propaganda with computational methods, and 3) studying the discourse surrounding science preprints (i.e., papers that have been posted to the Internet but have not been peer reviewed). The program culminated in an exposition where each team presented research results to program officers, senior faculty, deans, government officials, and industry experts. Here we present the program curriculum, metrics of educational effectiveness, and feedback collected from participants. © American Society for Engineering Education, 2022.

Quantitative PBMC ADCC bioassay and ADCC reporter bioassays for immunotherapy and SARS-CoV-2 monoclonal antibody development

Fc effector function is one of the main mechanisms of action (MoA) for therapeutic monoclonal antibodies (mAbs). Quantitative measurement of antibody-dependent cellular cytotoxicity (ADCC) is critically required for understanding the Fc function in mAb drug development. Despite the increasing interest and clinical success of the mAb therapeutic, it has been highly challenging to measure their ADCC activity in a reproducible and quantitative manner due to the lack of consistency in current methods that are based on primary PBMCs or NK cells and use tedious assay procedures. To improve ADCC assay precision so they can be validated as potency assay in cGMP laboratories, we developed reporter based ADCC bioassays using engineered effector cell line stably expressing a luciferase reporter and FcγRIIIa (V or F variant) to replace primary PBMC to overcome the assay variation. The ADCC reporter bioassays have been validated according to ICH guidelines by many laboratories and are demonstrated to be suitable for product release and stability studies in a quality-controlled environment. For early research and antibody characterization, we developed an improved PBMC ADCC assay using ADCC-prequalified PBMCs and engineered HiBiT target cells so they can measure the target specific lysis in ADCC. The PBMCs used in the study are isolated from prescreened blood donors and QC tested in ADCC assay. When HiBiT target cells are incubated with an antibody and PBMCs, HiBiT are released to the culture medium where it binds to LgBiT in the detection reagent to form a functional NanoBiT luciferase to generate luminescence signal. This new PBMC ADCC bioassay is simple, homogenous, highly sensitive, and gives a robust assay window. We demonstrate that it can quantitatively measure the potency for mAb drugs in cancer immunotherapy (e.g., rituximab, trastuzumab), and for anti-SARS-CoV-2 spike antibodies in antiviral drug development. Additionally, it shows antibody potency comparable with the ADCC reporter bioassay. In summary, the new PBMC ADCC bioassay using HiBiT target cells can be a valuable tool for early antibody discovery and characterization and also for method bridging study with ADCC reporter bioassay.

Early prediction model for disease progression of COV1D-19 patients based on XGBoost: Establishment and evaluation

Objective To construct an XGBoost prediction model to predict disease severity of COVID-19 based on clinical characteristics dataset of COVID-19 patients.Methods A total of 347 laboratory-confirmed COVID-19 patients with complete medical information admitted from Feb 10 to April 5, 2020 were screened from the medical record system of Huoshenshan Hospital.Firstly, 21 features with significant differences were screened out as input features for the training model.Bayesian optimization was performed on the constructed XGBoost model to adjust the parameters, and the optimal combination of features was filtered based on feature importance.To further analyze the positive and negative effects of the numerical size of each feature on the prediction results, each feature importance was quantified and attributed by using SHapley Additive explanations (SHAP).Finally, the performance of the XGBoost prediction model was evaluated, and the model was compared and discussed with other machine learning methods, including support vector machine (SVM), naive Bayes ( NB ) , logical regression ( LR) , and k-nearest neighbors ( KNN ).Results In this study, 21 features with significant differences between the severe and non-severe groups were selected for training and validation.The optimal subset with 10 features in the k-nearest neighbor model obtained the highest value of area under curve ( AUG) among the 4 models in the validation set.XGBoost and support vector machine were better than other machine learning methods in terms of prediction performance (AUG;0.942 0, and 0.959 4 on the test set, respectively) , and the training speed of XGBoost was significantly faster.Conclusion A prediction model based on XGBoost is successfully built to achieve early prediction of disease severity of GOVID-19 patients. © 2022 Journal of Army Medical University. All rights reserved.

Study on the Correlation Between Indoor Light Environment and Human Fatigue

Due to the continuing impact of COVID-19, people spend an increasing amount of time on working or studying from home. The indoor light environment became quite important since it can affect people's physical and mental health. In order to reduce human fatigue resulting from continuous indoor working or studying and to improve work efficiency, and also hope to contribute to the research of indoor light environment design, this paper explores the correlation between indoor light environments and fatigue. Through laboratory simulation of indoor light environment. Participants are asked to complete the task stimulation test and filled in the subjective fatigue questionnaire with three different illumination levels. Their EEG (ElectroEncephaloGraphy), eye movement, and other physiological data are also monitored at the same time. The participants’ fatigue degree is statistically analyzed under the 300 lx, 570 lx, and 870 lx illumination. The results showed that the lighting environment has a strong correlation with human fatigue. Fatigue degree varies the most from 570 to 870 lx. There is a largest error rate gap of task test up to 20% under 300 lx and 870 lx illumination. As the illuminance increases, the fatigue degree has a visible trend of increase as well, and it is the most obvious under the 870 lx illumination. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Restaurant employees' attitudinal reactions to social distancing difficulties: a multi-wave study

Purpose: Grounded in the job demands–resources (JD-R) theory, this study investigates how the difficulty in social distancing at work, resulting from the COVID-19 crisis, may lead to intention to quit and career regret and how and when these effects may be attenuated. Design/methodology/approach: Three-wave survey data were collected from 223 frontline service workers in a large restaurant company during the COVID-19 crisis. Findings: The results show that difficulty in social distancing reduced employees' work engagement, and consequently, increased their turnover intention and career regret. These relationships were moderated by external employability, such that the influence of difficulty in social distancing weakened as external employability increased. Originality/value: Social distancing measures have been applied across the globe to minimize transmission of COVID-19. However, such measures create a new job demand for service workers who find it difficult to practice social distancing due to the high contact intensity of service delivery. This study identified personal resources that help service workers cope with the demand triggered by COVID-19. © 2022, Emerald Publishing Limited.

Traditional Chinese medicine together with high-dose vitamin C improves the therapeutic effect of western medicine against COVID-19

Objectives: Traditional Chinese medicine has been reported to be effective in the treatment of epidemic diseases. Here, we aimed to investigate the effects of combined therapy of Chinese and western medicine on coronavirus disease 2019 (COVID-19). Methods: A total of 60 patients diagnosed with COVID-19 were enrolled. Both the ordinary and severely affected patients were randomly divided into Groups A-C each with 10 cases each. The patients in Group A-C received Western medicine, Western medicine + traditional Chinese medicine, and Western medicine + traditional Chinese medicine + high dose of vitamin C, respectively. The time of disease recovery, symptoms disappearance, chest CT improvement, and tongue amelioration was recorded. Leukocyte, neutrophil and lymphocyte were monitored, as well as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), procalitonin (PCT), inflammatory factors, partial pressure of oxygen and carbon dioxide (PaCO2) and oxygenation index (PaO2). Urinary tract stones, liver function, and other side-effects such as gastrointestinal dysfunction were also investigated. Results: Traditional Chinese medicine enhanced the effect of Western medicine, including the reduction of CRP, ESR, PCT, and inflammatory factors, and the increase of leukocyte, neutrophil, and lymphocyte counts, and the improvement of respiratory rate, PaO2, PaCO2, and oxygenation index. Traditional Chinese medicine combined with high-dose Vitamin C therapy more effectively shortened the time of disease recovery, symptom disappearance, chest CT improvement, and tongue amelioration. Conclusions: a combined therapy of Western medicine, traditional Chinese medicine, and high dose of Vitamin C results in a most effective outcome in the treatment of COVID-19.

Isolation and management measures in hospitals during the COVID⁃19 epidemic

Prevention and control measures for isolation and centralized management of COVID‑19 patients and suspected patients are key measures in the COVID‑19 pandemic. It is crucial to rapidly activate the emergency plan and establish an isolation ward if confirmed or suspect COVID‑19 cases are found. Patients and staff management policies and procedures should be adopted and implemented. Strict implementation of the disinfection and isolation system is very important. Personal protection of patients and medical staff, and standardized disinfection of the ward should be implemented. Personalized treatment plan should be made according to the severity of each patient's condition, the situation of chronic diseases and the severity of COVID‑19. Great attention, strictly implementing the disinfection and isolation system, and standardizing the management of specific work can effectively help controling the infection in hospital. © 2022, Editorial Board of Medical Journal of Wuhan University. All right reserved.

Clinical characteristics of discharged COVID‑19 patients with reappeared positive nucleic acid test

Objective: To investigate the incidence of reappearance of positive results of nucleic acid (NA) test in discharged COVID‑19 patients, and retrospectively evaluate and compare the clinical characteristics between reappeared NA‑positive group and NA‑negative group, so as to provide the decision‑making basis for the quarantine management, health monitoring, and epidemic prevention. Methods: A total of 188 cases of discharged COVID‑19 patients from 3 quarantine observation sites in Wuhan from February to March 2020 were divided into reappeared NA‑positive group and NA‑negative group, based on the NA test. Clinical characteristics including general information, symptoms, underlying diseases, severity of disease, steroid administration, duration of hospitalization, duration from the beginning of positive to negative and time of onset of reappearance of positive for NA test, were compared between the two groups. Results: Of the 188 cases, 30 cases (15.96%) reappeared positive for NA test, and onset of reappearance ranged from day 1 to day 16 since the last NA negative, with a median time of 5 days after onset. A total of 51 cases (32.28%) in NA‑negative group had underlying diseases, while 18 (60.0%) cases reappeared NA‑positive had underlying diseases, and the rate of underlying diseases showed significant difference between the two groups (P<0.01). However, there were no statistical differences in gender, age, and symptoms on admission, severity of disease, steroid administration, duration from the beginning of positive to negative for NA test between the two groups. Conclusion: Since COVID‑19 patients with underlying diseases showed higher incidence of reappearance of positive NA test, COVID‑19 patients after discharge need to strengthen for quarantine control, further continuous health monitoring, and re‑tests of SARS‑CoV‑2 nucleic acid. © 2021, Editorial Board of Medical Journal of Wuhan University. All right reserved.

Short-Term Power Load Forecasting Method Based on FI under the Impact of Epidemic

The sudden COVID-19 epidemic has caused a serious impact on power load. In order to effectively deal with the impact of the epidemic and improve the accuracy of short-term load prediction under the impact of the epidemic, a short-term power load forecasting method based on fear index (FI) under the impact of epidemic was proposed. Firstly, epidemic data was used to construct the FI, together with the time information, historical load and meteorological conditions as the input variables of generalized regression neural network (GRNN) model. And then a fruit fly optimization algorithm (FOA) was used to optimize the GRNN smoothing factor to improve the accuracy and stability of the predicted results. Finally, the model was used to make the prediction. The simulation results show that this method can effectively improve the accuracy of short-term load forecasting under the impact of epidemic and provide reference for short-term load forecasting under the impact of major disasters. © 2021, Editorial Department of Transaction of Beijing Institute of Technology. All right reserved.

Research on public risk perception of the public health emergency in remission period

The remission period of public health emergencies is the last mile to solve the problem, and is of great significance throughout the life cycle of public health emergencies. The level of public risk perception is of great significance to the implementation of epidemic prevention policies. However, at present, there are few studies on public risk perception of public health emergencies, and there is no research on public risk perception based on the development stage of the event. Based on this, this article takes the case of the COVID-19 as an example, combs the research results of public health emergencies at home and abroad, and uses the KAB model (Knowledge-Attitude-Behavior) as the theoretical basis to establish an index system for public risk perception evaluation of public health emergencies during the remission period, and use a questionnaire survey to conduct an empirical study on the degree of public risk perception of the public health emergencies during the remission period. © 2020 IEEE.

Production Resumption Analysis for Different Industries by K-means Clustering

At the beginning of 2020, the COVID-19 suddenly struck, which had a great impact on industrial production, causing most companies to shut down. To further study the work and production resumption progress for industrial users, this paper analyzes the electricity consumption data from a total of 200 users in four industries including heavy industries, general industry and commerce, agricultural production industries, and agricultural irrigation and drainage industries in a city in Eastern China before and after the Spring Festival and pandemic. K-means clustering with Silhouette Coefficient is adopted as the primary tool to analyze the resumption thresholds for different industries. Combined with these thresholds, the industry resumption rate curve is further drawn, and the resumption rate of each industry one month before and after the Spring Festival is analyzed to provide support for in-depth analysis of the impact of the pandemic on different industries. © 2020 IEEE.