Trick or Drink: Offline and Social Media Hierarchical Normative Influences on Halloween Celebration Drinking.

The current study examined the relationship between perceived societal and personal celebration drinking norms, social media use, and alcohol consumption during Halloween. The study used a survey of a nationally representative, convenience, and cross-sectional sample of underage youth (18-20 years old; N = 525). Participants self-reported their own drinking, perceived descriptive norms among peers and close friends, and alcohol-related social media posting and interaction during Halloween. Results revealed that underage youth's estimation of societal drinking norms related to their proximal close friends' drinking norms, which in turn, influenced self-reported number of drinks consumed during Halloween. Social media posting and interaction with alcohol-related content were associated with greater descriptive normative perceptions and self-reported drinking. Extending the hierarchical social norms approach, our findings showed that normative perceptions about proximal reference groups' drinking, along with alcohol-related social media activities, were associated with greater number of drinks consumed during Halloween.

Ecological prediction at macroscales using big data: Does sampling design matter?

Although ecosystems respond to global change at regional to continental scales (i.e., macroscales), model predictions of ecosystem responses often rely on data from targeted monitoring of a small proportion of sampled ecosystems within a particular geographic area. In this study, we examined how the sampling strategy used to collect data for such models influences predictive performance. We subsampled a large and spatially extensive data set to investigate how macroscale sampling strategy affects prediction of ecosystem characteristics in 6,784 lakes across a 1.8-million-km2 area. We estimated model predictive performance for different subsets of the data set to mimic three common sampling strategies for collecting observations of ecosystem characteristics: random sampling design, stratified random sampling design, and targeted sampling. We found that sampling strategy influenced model predictive performance such that (1) stratified random sampling designs did not improve predictive performance compared to simple random sampling designs and (2) although one of the scenarios that mimicked targeted (non-random) sampling had the poorest performing predictive models, the other targeted sampling scenarios resulted in models with similar predictive performance to that of the random sampling scenarios. Our results suggest that although potential biases in data sets from some forms of targeted sampling may limit predictive performance, compiling existing spatially extensive data sets can result in models with good predictive performance that may inform a wide range of science questions and policy goals related to global change.

Celebration Drinking around the Clock.

The fact that St. Patrick's Day (SPD) celebration drinking occurs during a specified, public, and socially-acceptable time frame which spans the better part of a day and evening makes it an important time to understand and attempt to influence celebration drinking behaviors among young adults. SPD has been identified as the celebration during which college students consume more alcohol than any other point during the school year. Intervention opportunities can be more successful with an understanding of the factors associated with alcohol consumption at specific times on particular celebrations. This study examined the factors associated with celebration drinking at different time periods on SPD which included perceived descriptive and injunctive norms, the numbers of close friends and acquaintances present, social media relationships, demographic variables, past drinking behavior, and intent to drink on SPD at the three time points of interest. Findings showed variability in the predictive factors on SPD celebration drinking at different times of the day. The theoretical and practical intervention implications of the findings are discussed.

Vaccine rhetoric on social media and COVID-19 vaccine uptake rates: A triangulation using self-reported vaccine acceptance.

The primary goal of this study is to examine the association between vaccine rhetoric on Twitter and the public's uptake rates of COVID-19 vaccines in the United States, compared to the extent of an association between self-reported vaccine acceptance and the CDC's uptake rates. We downloaded vaccine-related posts on Twitter in real-time daily for 13 months, from October 2021 to September 2022, collecting over half a billion tweets. A previously validated deep-learning algorithm was then applied to (1) filter out irrelevant tweets and (2) group the remaining relevant tweets into pro-, anti-, and neutral vaccine sentiments. Our results indicate that the tweet counts (combining all three sentiments) were significantly correlated with the uptake rates of all stages of COVID-19 shots (p < 0.01). The self-reported level of vaccine acceptance was not correlated with any of the stages of COVID-19 shots (p > 0.05) but with the daily new infection counts. These results suggest that although social media posts on vaccines may not represent the public's opinions, they are aligned with the public's behaviors of accepting vaccines, which is an essential step for developing interventions to increase the uptake rates. In contrast, self-reported vaccine acceptance represents the public's opinions, but these were not correlated with the behaviors of accepting vaccines. These outcomes provide empirical support for the validity of social media analytics for gauging the public's vaccination behaviors and understanding a nuanced perspective of the public's vaccine sentiment for health emergencies.

Unsupervised Anomaly Detection by Robust Density Estimation.

Density estimation is a widely used method for unsupervised anomaly detection. However, the presence of anomalies in training data may severely impact the density estimation process, thereby hampering the use of more sophisticated density estimation methods such as those based on deep neural networks. In this work, we propose RobustRealNVP, a robust deep density estimation framework for unsupervised anomaly detection. Our approach differs from existing flow-based models from two perspectives. First, RobustRealNVP discards data points with low estimated densities during optimization to prevent them from corrupting the density estimation process. Furthermore, it imposes Lipschitz regularization to ensure smoothness in the estimated density function. We demonstrate the robustness of our algorithm against anomalies in training data from both theoretical and empirical perspectives. The results show that our algorithm outperforms state-of-the-art unsupervised anomaly detection methods.

Using Deep Learning to Identify Linguistic Features that Facilitate or Inhibit the Propagation of Anti- and Pro-Vaccine Content on Social Media.

Anti-vaccine content is rapidly propagated via social media, fostering vaccine hesitancy, while pro-vaccine content has not replicated the opponent's successes. Despite this disparity in the dissemination of anti- and pro-vaccine posts, linguistic features that facilitate or inhibit the propagation of vaccine-related content remain less known. Moreover, most prior machine-learning algorithms classified social-media posts into binary categories (e.g., misinformation or not) and have rarely tackled a higher-order classification task based on divergent perspectives about vaccines (e.g., anti-vaccine, pro-vaccine, and neutral). Our objectives are (1) to identify sets of linguistic features that facilitate and inhibit the propagation of vaccine-related content and (2) to compare whether anti-vaccine, provaccine, and neutral tweets contain either set more frequently than the others. To achieve these goals, we collected a large set of social media posts (over 120 million tweets) between Nov. 15 and Dec. 15, 2021, coinciding with the Omicron variant surge. A two-stage framework was developed using a fine-tuned BERT classifier, demonstrating over 99 and 80 percent accuracy for binary and ternary classification. Finally, the Linguistic Inquiry Word Count text analysis tool was used to count linguistic features in each classified tweet. Our regression results show that anti-vaccine tweets are propagated (i.e., retweeted), while pro-vaccine tweets garner passive endorsements (i.e., favorited). Our results also yielded the two sets of linguistic features as facilitators and inhibitors of the propagation of vaccine-related tweets. Finally, our regression results show that anti-vaccine tweets tend to use the facilitators, while pro-vaccine counterparts employ the inhibitors. These findings and algorithms from this study will aid public health officials' efforts to counteract vaccine misinformation, thereby facilitating the delivery of preventive measures during pandemics and epidemics.

Learning Deep Neural Networks under Agnostic Corrupted Supervision.

Training deep neural models in the presence of corrupted supervision is challenging as the corrupted data points may significantly impact the generalization performance. To alleviate this problem, we present an efficient robust algorithm that achieves strong guarantees without any assumption on the type of corruption, and provides a unified framework for both classification and regression problems. Unlike many existing approaches that quantify the quality of the data points (e.g., based on their individual loss values), and filter them accordingly, the proposed algorithm focuses on controlling the collective impact of data points on the average gradient. Even when a corrupted data point failed to be excluded by our algorithm, the data point will have very limited impact on the overall loss, as compared with state-of-the-art filtering methods based on loss values. Extensive experiments on multiple benchmark datasets have demonstrated the robustness of our algorithm under different types of corruptions.

RCA: A Deep Collaborative Autoencoder Approach for Anomaly Detection.

Unsupervised anomaly detection (AD) plays a crucial role in many critical applications. Driven by the success of deep learning, recent years have witnessed growing interest in applying deep neural networks (DNNs) to AD problems. A common approach is using autoencoders to learn a feature representation for the normal observations in the data. The reconstruction error of the autoencoder is then used as outlier score to detect the anomalies. However, due to the high complexity brought upon by over-parameterization of DNNs, the reconstruction error of the anomalies could also be small, which hampers the effectiveness of these methods. To alleviate this problem, we propose a robust framework using collaborative autoencoders to jointly identify normal observations from the data while learning its feature representation. We investigate the theoretical properties of the framework and empirically show its outstanding performance as compared to other DNN-based methods. Empirical results also show resiliency of the framework to missing values compared to other baseline methods.

Automated Analysis of the US Drought Monitor Maps With Machine Learning and Multiple Drought Indicators.

The US Drought Monitor (USDM) is a hallmark in real time drought monitoring and assessment as it was developed by multiple agencies to provide an accurate and timely assessment of drought conditions in the US on a weekly basis. The map is built based on multiple physical indicators as well as reported observations from local contributors before human analysts combine the information and produce the drought map using their best judgement. Since human subjectivity is included in the production of the USDM maps, it is not an entirely clear quantitative procedure for other entities to reproduce the maps. In this study, we developed a framework to automatically generate the maps through a machine learning approach by predicting the drought categories across the domain of study. A persistence model served as the baseline model for comparison in the framework. Three machine learning algorithms, logistic regression, random forests, and support vector machines, with four different groups of input data, which formed an overall of 12 different configurations, were used for the prediction of drought categories. Finally, all the configurations were evaluated against the baseline model to select the best performing option. The results showed that our proposed framework could reproduce the drought maps to a near-perfect level with the support vector machines algorithm and the group 4 data. The rest of the findings of this study can be highlighted as: 1) employing the past week drought data as a predictor in the models played an important role in achieving high prediction scores, 2) the nonlinear models, random forest, and support vector machines had a better overall performance compared to the logistic regression models, and 3) with borrowing the neighboring grid cells information, we could compensate the lack of training data in the grid cells with insufficient historical USDM data particularly for extreme and exceptional drought conditions.

Using Machine Learning to Compare Provaccine and Antivaccine Discourse Among the Public on Social Media: Algorithm Development Study.

BACKGROUND: Despite numerous counteracting efforts, antivaccine content linked to delays and refusals to vaccinate has grown persistently on social media, while only a few provaccine campaigns have succeeded in engaging with or persuading the public to accept immunization. Many prior studies have associated the diversity of topics discussed by antivaccine advocates with the public's higher engagement with such content. Nonetheless, a comprehensive comparison of discursive topics in pro- and antivaccine content in the engagement-persuasion spectrum remains unexplored. OBJECTIVE: We aimed to compare discursive topics chosen by pro- and antivaccine advocates in their attempts to influence the public to accept or reject immunization in the engagement-persuasion spectrum. Our overall objective was pursued through three specific aims as follows: (1) we classified vaccine-related tweets into provaccine, antivaccine, and neutral categories; (2) we extracted and visualized discursive topics from these tweets to explain disparities in engagement between pro- and antivaccine content; and (3) we identified how those topics frame vaccines using Entman's four framing dimensions. METHODS: We adopted a multimethod approach to analyze discursive topics in the vaccine debate on public social media sites. Our approach combined (1) large-scale balanced data collection from a public social media site (ie, 39,962 tweets from Twitter); (2) the development of a supervised classification algorithm for categorizing tweets into provaccine, antivaccine, and neutral groups; (3) the application of an unsupervised clustering algorithm for identifying prominent topics discussed on both sides; and (4) a multistep qualitative content analysis for identifying the prominent discursive topics and how vaccines are framed in these topics. In so doing, we alleviated methodological challenges that have hindered previous analyses of pro- and antivaccine discursive topics. RESULTS: Our results indicated that antivaccine topics have greater intertopic distinctiveness (ie, the degree to which discursive topics are distinct from one another) than their provaccine counterparts (t122=2.30, P=.02). In addition, while antivaccine advocates use all four message frames known to make narratives persuasive and influential, provaccine advocates have neglected having a clear problem statement. CONCLUSIONS: Based on our results, we attribute higher engagement among antivaccine advocates to the distinctiveness of the topics they discuss, and we ascribe the influence of the vaccine debate on uptake rates to the comprehensiveness of the message frames. These results show the urgency of developing clear problem statements for provaccine content to counteract the negative impact of antivaccine content on uptake rates.

Improving Heart Disease Risk Through Quality-Focused Diet Logging: Pre-Post Study of a Diet Quality Tracking App.

BACKGROUND: Diet-tracking mobile apps have gained increased interest from both academic and clinical fields. However, quantity-focused diet tracking (eg, calorie counting) can be time-consuming and tedious, leading to unsustained adoption. Diet quality-focusing on high-quality dietary patterns rather than quantifying diet into calories-has shown effectiveness in improving heart disease risk. The Healthy Heart Score (HHS) predicts 20-year cardiovascular risks based on the consumption of foods from quality-focused food categories, rather than detailed serving sizes. No studies have examined how mobile health (mHealth) apps focusing on diet quality can bring promising results in health outcomes and ease of adoption. OBJECTIVE: This study aims to design a mobile app to support the HHS-informed quality-focused dietary approach by enabling users to log simplified diet quality and view its real-time impact on future heart disease risks. Users were asked to log food categories that are the main predictors of the HHS. We measured the app's feasibility and efficacy in improving individuals' clinical and behavioral factors that affect future heart disease risks and app use. METHODS: We recruited 38 participants who were overweight or obese with high heart disease risk and who used the app for 5 weeks and measured weight, blood sugar, blood pressure, HHS, and diet score (DS)-the measurement for diet quality-at baseline and week 5 of the intervention. RESULTS: Most participants (30/38, 79%) used the app every week and showed significant improvements in DS (baseline: mean 1.31, SD 1.14; week 5: mean 2.36, SD 2.48; 2-tailed t test t29=-2.85; P=.008) and HHS (baseline: mean 22.94, SD 18.86; week 4: mean 22.15, SD 18.58; t29=2.41; P=.02) at week 5, although only 10 participants (10/38, 26%) checked their HHS risk scores more than once. Other outcomes, including weight, blood sugar, and blood pressure, did not show significant changes. CONCLUSIONS: Our study showed that our logging tool significantly improved dietary choices. Participants were not interested in seeing the HHS and perceived logging diet categories irrelevant to improving the HHS as important. We discuss the complexities of addressing health risks and quantity- versus quality-based health monitoring and incorporating secondary behavior change goals that matter to users when designing mHealth apps.

Diffusion size and structural virality: The effects of message and network features on spreading health information on twitter.

Relying on diffusion of innovation theory, this study examines the impacts of perceived message features and network characteristics on size (i.e., the number of retweets a message receives) and structural virality (i.e., quantified distinction between broadcast and viral diffusion) of information diffusion on Twitter. The study collected 425 unique tweets posted by CDC during a 17-week period and constructed a diffusion tree for each unique tweet. Findings indicated that, with respect to message features, perceived efficacy after reading a tweet positively predicted diffusion size of the tweet, whereas perceived susceptibility to a health condition after reading a tweet positively predicted structural virality of the tweet. Perceived negative emotion positively predicted both size and structural virality. With respect to network features, the level of involvement of brokers in diffusing a tweet increased the tweet's structural virality. Theoretical and practical implications were discussed on disseminating health information via broadcasting and viral diffusion on social media.

Creating multithemed ecological regions for macroscale ecology: Testing a flexible, repeatable, and accessible clustering method.

Understanding broad-scale ecological patterns and processes often involves accounting for regional-scale heterogeneity. A common way to do so is to include ecological regions in sampling schemes and empirical models. However, most existing ecological regions were developed for specific purposes, using a limited set of geospatial features and irreproducible methods. Our study purpose was to: (1) describe a method that takes advantage of recent computational advances and increased availability of regional and global data sets to create customizable and reproducible ecological regions, (2) make this algorithm available for use and modification by others studying different ecosystems, variables of interest, study extents, and macroscale ecology research questions, and (3) demonstrate the power of this approach for the research question-How well do these regions capture regional-scale variation in lake water quality? To achieve our purpose we: (1) used a spatially constrained spectral clustering algorithm that balances geospatial homogeneity and region contiguity to create ecological regions using multiple terrestrial, climatic, and freshwater geospatial data for 17 northeastern U.S. states (~1,800,000 km2); (2) identified which of the 52 geospatial features were most influential in creating the resulting 100 regions; and (3) tested the ability of these ecological regions to capture regional variation in water nutrients and clarity for ~6,000 lakes. We found that: (1) a combination of terrestrial, climatic, and freshwater geospatial features influenced region creation, suggesting that the oft-ignored freshwater landscape provides novel information on landscape variability not captured by traditionally used climate and terrestrial metrics; and (2) the delineated regions captured macroscale heterogeneity in ecosystem properties not included in region delineation-approximately 40% of the variation in total phosphorus and water clarity among lakes was at the regional scale. Our results demonstrate the usefulness of this method for creating customizable and reproducible regions for research and management applications.

Building a multi-scaled geospatial temporal ecology database from disparate data sources: fostering open science and data reuse.

Although there are considerable site-based data for individual or groups of ecosystems, these datasets are widely scattered, have different data formats and conventions, and often have limited accessibility. At the broader scale, national datasets exist for a large number of geospatial features of land, water, and air that are needed to fully understand variation among these ecosystems. However, such datasets originate from different sources and have different spatial and temporal resolutions. By taking an open-science perspective and by combining site-based ecosystem datasets and national geospatial datasets, science gains the ability to ask important research questions related to grand environmental challenges that operate at broad scales. Documentation of such complicated database integration efforts, through peer-reviewed papers, is recommended to foster reproducibility and future use of the integrated database. Here, we describe the major steps, challenges, and considerations in building an integrated database of lake ecosystems, called LAGOS (LAke multi-scaled GeOSpatial and temporal database), that was developed at the sub-continental study extent of 17 US states (1,800,000 km(2)). LAGOS includes two modules: LAGOSGEO, with geospatial data on every lake with surface area larger than 4 ha in the study extent (~50,000 lakes), including climate, atmospheric deposition, land use/cover, hydrology, geology, and topography measured across a range of spatial and temporal extents; and LAGOSLIMNO, with lake water quality data compiled from ~100 individual datasets for a subset of lakes in the study extent (~10,000 lakes). Procedures for the integration of datasets included: creating a flexible database design; authoring and integrating metadata; documenting data provenance; quantifying spatial measures of geographic data; quality-controlling integrated and derived data; and extensively documenting the database. Our procedures make a large, complex, and integrated database reproducible and extensible, allowing users to ask new research questions with the existing database or through the addition of new data. The largest challenge of this task was the heterogeneity of the data, formats, and metadata. Many steps of data integration need manual input from experts in diverse fields, requiring close collaboration.