Search Query Data Research Articles

Public and scholarly interest in social robots: An investigation through Google Trends, bibliometric analysis, and systematic literature review

The COVID-19 pandemic has expedited the integration of social robots within the healthcare sector. This research employs a tripartite methodology, combining Google Trends analysis, bibliometric analysis, and a systematic literature review, to gauge both public and research interest in social robots within the healthcare domain. In the Google Trends analysis, search query data for “Social Robots” was retrieved, encompassing both “all categories” and the specific “health” category. Seasonal effects on relative search volumes (RSV) were assessed through the cosinor model. The analysis confirmed statistically significant seasonal patterns in RSV for “social robots” within the “health” category. Conversely, for the broader “all categories,” only the intercept showed significance, while sinw and cosw were deemed insignificant. For bibliometric analysis, the global literature on “robotics” and “healthcare” was examined in the SCOPUS database. From the extensive pool of publications, 144 relevant studies were identified out of 4037 publications. These studies were further analyzed using VOSviewer, providing insights into recent trends and hot topics concerning social robots in healthcare. The systematic literature review focused on 46 articles published from 2019 to the end of 2023. The findings revealed a lack of consensus on the drivers, barriers, and outcomes associated with social robot acceptance and human-robot interaction (HRI). The study systematically maps the existing research on these aspects, introducing a novel categorization and presenting the concept of a “robot user's ecosystem.” This concept emphasizes the imperative involvement of all stakeholders in the development and understanding of social robots. Ultimately, this methodological approach not only identifies nine research gaps in the current literature but also formulates numerous research questions to guide future researchers in this domain.

Optimizing demand forecasting for business events tourism: a comparative analysis of cutting-edge models

PurposeGiven the unique nature of business events tourism, this paper evaluates the forecasting performance of various models using search query data (SQD) to forecast convention attendance.Design/methodology/approachThis research uses monthly and quarterly business event attendance data from both the U.S. (Las Vegas) and China (Macau) markets. Using SQD as the input, we evaluated and compared the cutting-edge forecasting models including Prophet and Long Short-Term Memory (LSTM).FindingsThe study reveals that Prophet outperforms complex neural network models in forecasting business event tourism demand. Keywords related to convention facilities, conventions or exhibitions, and transportation are proven to be useful in forecasting business travel demand.Practical implicationsProphet is an accessible forecasting model for event-tourism practitioners, especially useful in the volatile business event tourism sector. Using verified search keywords in models helps understand traveler motivations and aids event planning.Originality/valueOur study is among the first to empirically evaluate the performance of forecasting models for business travel demand. In comparison with other mainstream forecasting models, our study extends the scope to examine both the U.S. and Chinese markets.

The Interest in Betting, Smoking, Alcohol, and Drugs in Malawi: Changing Trends between 2012-2022.

In recent years, the online gambling market has rapidly developed, and betting has become one of the most popular forms of gambling. The aim of this study was to analyse the interest of the Malawian population in terms related to betting, sports betting, alcohol, cigarettes, and some psychoactive drugs through the relative search volumes of Google Trends. Internet search query data related to betting, sports betting, alcohol, cigarettes, and psychoactive drugs were obtained monthly from Google Trends for the period 2010-2022. Comparisons of interest levels in these topics were conducted in Malawi, and correlation coefficients were calculated. In Malawi, relative search volumes for betting and sports betting terms were the highest (average RSVs: 66% and 30%). It was found that from 2019 onwards, the interest in betting and sports-related search topics and keywords increased significantly (p < 0.001). Strong positive correlations were found between betting-related keywords and alcohol and gross domestic product (r = 0.831 and r = 0.901, p < 0.001). A positive correlation was found between betting and psychoactive drug-related terms (minimum r = 0.417, p < 0.01). This study concludes that the interest of the Malawian population in betting has increased in recent years, while interest in psychoactive drugs and alcohol remains high. Gross domestic product is highly correlated with society's interest in betting. It was additionally found that Google Trends can be used as a tool to predict and monitor future risky behaviours, such as gambling disorder.

A Novel Framework to Forecast COVID-19 Incidence Based on Google Trends Search Data

The global outbreak of coronavirus disease 2019 (COVID-19) has spread to more than 200 countries worldwide, leading to severe health and socioeconomic consequences. As such, the topic of monitoring and predicting epidemics has been attracting a lot of interest. Previous work reported search volumes from Google Trends are beneficial in decoding influenza dynamics, implying its potential for COVID-19 prediction. Therefore, a predictive model using the Wiener methods was built based on epidemic-related search queries from Google Trends, along with climate variables, aiming to forecast the dynamics of the weekly COVID-19 incidence in Washington, DC, USA. The Wiener model, which shares the merits of interpretability, low computation costs, and adaptation to nonlinear fluctuations, was used in this study. Models with multiple sets of features were constructed and further optimized by the highest weight selecting strategy. Furthermore, comparisons to the other two commonly used prediction models based on the autoregressive integrated moving average (ARIMA) and long short-term memory (LSTM) were also performed. Our results showed the predicted COVID-19 trends significantly correlated with the actual (rho <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$=$</tex-math> </inline-formula> 0.88, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$p $</tex-math> </inline-formula> <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$&lt;$</tex-math> </inline-formula> 0.0001), outperforming those with ARIMA and LSTM approaches, indicating Google Trends data as a useful tool in terms of COVID-19 prediction. Also, the model using 20 search queries with the highest weighting outperformed all other models, supporting the highest weight feature selection as a feasible criterion. Google Trends search query data can be used to forecast the outbreak of COVID-19, which might assist health policymakers to allocate health care resources and taking preventive strategies.

Straight to the point: evaluation of a Point of Care Information (POCI) resource in answering disease-related questions.

To evaluate the ability of DynaMedex, an evidence-based drug and disease Point of Care Information (POCI) resource, in answering clinical queries using keyword searches. Real-world disease-related questions compiled from clinicians at an academic medical center, DynaMedex search query data, and medical board review resources were categorized into five clinical categories (complications & prognosis, diagnosis & clinical presentation, epidemiology, prevention & screening/monitoring, and treatment) and six specialties (cardiology, endocrinology, hematology-oncology, infectious disease, internal medicine, and neurology). A total of 265 disease-related questions were evaluated by pharmacist reviewers based on if an answer was found (yes, no), whether the answer was relevant (yes, no), difficulty in finding the answer (easy, not easy), cited best evidence available (yes, no), clinical practice guidelines included (yes, no), and level of detail provided (detailed, limited details). An answer was found for 259/265 questions (98%). Both reviewers found an answer for 241 questions (91%), neither found the answer for 6 questions (2%), and only one reviewer found an answer for 18 questions (7%). Both reviewers found a relevant answer 97% of the time when an answer was found. Of all relevant answers found, 68% were easy to find, 97% cited best quality of evidence available, 72% included clinical guidelines, and 95% were detailed. Recommendations for areas of resource improvement were identified. The resource enabled reviewers to answer most questions easily with the best quality of evidence available, providing detailed answers and clinical guidelines, with a high level of replication of results across users.

Применение Google Trends для прогнозирования миграции из России: агрегация поисковых запросов и учет лаговой структуры

This paper proposes an approach for predicting migration statistics using Google Trends Index (GTI) search query data. We improved the existing methodology in two directions: firstly, we proposed an approach of aggregating key search queries based on various statistical criteria; secondly, we showed the importance of including in the migration model the time lag structure of search queries, depending on migration goals and the associated GTIs. We demonstrate the performance of the proposed approaches on monthly data from the German statistical office on migration volume from Russia to Germany from January 2011 to August 2022. The results show that distributed lag migration models with GTI are better predict migration than SARIMA models. Average lag estimates, i.e. the reaction time of migration statistics to search queries on the topics “embassy”, “work” and “study”, were 5.6, 6.5 and 8 months, respectively. We demonstrate that for forecasting migration from Russia to Germany, it is sufficient to consider only search queries related to the topic “embassy”.

The unintended consequences of inconsistent closure policies and mobility restrictions during epidemics

BackgroundControlling the spread of infectious diseases―even when safe, transmission-blocking vaccines are available―may require the effective use of non-pharmaceutical interventions (NPIs), e.g., mask wearing, testing, limits on group sizes, venue closure. During the SARS-CoV-2 pandemic, many countries implemented NPIs inconsistently in space and time. This inconsistency was especially pronounced for policies in the United States of America (US) related to venue closure.MethodsHere, we investigate the impact of inconsistent policies associated with venue closure using mathematical modeling and high-resolution human mobility, Google search, and county-level SARS-CoV-2 incidence data from the USA. Specifically, we look at high-resolution location data and perform a US-county-level analysis of nearly 8 million SARS-CoV-2 cases and 150 million location visits, including 120 million church visitors across 184,677 churches, 14 million grocery visitors across 7662 grocery stores, and 13.5 million gym visitors across 5483 gyms.ResultsAnalyzing the interaction between venue closure and changing mobility using a mathematical model shows that, across a broad range of model parameters, inconsistent or partial closure can be worse in terms of disease transmission as compared to scenarios with no closures at all. Importantly, changes in mobility patterns due to epidemic control measures can lead to increase in the future number of cases. In the most severe cases, individuals traveling to neighboring jurisdictions with different closure policies can result in an outbreak that would otherwise have been contained. To motivate our mathematical models, we turn to mobility data and find that while stay-at-home orders and closures decreased contacts in most areas of the USA, some specific activities and venues saw an increase in attendance and an increase in the distance visitors traveled to attend. We support this finding using search query data, which clearly shows a shift in information seeking behavior concurrent with the changing mobility patterns.ConclusionsWhile coarse-grained observations are not sufficient to validate our models, taken together, they highlight the potential unintended consequences of inconsistent epidemic control policies related to venue closure and stress the importance of balancing the societal needs of a population with the risk of an outbreak growing into a large epidemic.

Influenza Epidemic Trend Surveillance and Prediction Based on Search Engine Data: Deep Learning Model Study.

Influenza outbreaks pose a significant threat to global public health. Traditional surveillance systems and simple algorithms often struggle to predict influenza outbreaks in an accurate and timely manner. Big data and modern technology have offered new modalities for disease surveillance and prediction. Influenza-like illness can serve as a valuable surveillance tool for emerging respiratory infectious diseases like influenza and COVID-19, especially when reported case data may not fully reflect the actual epidemic curve. This study aimed to develop a predictive model for influenza outbreaks by combining Baidu search query data with traditional virological surveillance data. The goal was to improve early detection and preparedness for influenza outbreaks in both northern and southern China, providing evidence for supplementing modern intelligence epidemic surveillance methods. We collected virological data from the National Influenza Surveillance Network and Baidu search query data from January 2011 to July 2018, totaling 3,691,865 and 1,563,361 respective samples. Relevant search terms related to influenza were identified and analyzed for their correlation with influenza-positive rates using Pearson correlation analysis. A distributed lag nonlinear model was used to assess the lag correlation of the search terms with influenza activity. Subsequently, a predictive model based on the gated recurrent unit and multiple attention mechanisms was developed to forecast the influenza-positive trend. This study revealed a high correlation between specific Baidu search terms and influenza-positive rates in both northern and southern China, except for 1 term. The search terms were categorized into 4 groups: essential facts on influenza, influenza symptoms, influenza treatment and medicine, and influenza prevention, all of which showed correlation with the influenza-positive rate. The influenza prevention and influenza symptom groups had a lag correlation of 1.4-3.2 and 5.0-8.0 days, respectively. The Baidu search terms could help predict the influenza-positive rate 14-22 days in advance in southern China but interfered with influenza surveillance in northern China. Complementing traditional disease surveillance systems with information from web-based data sources can aid in detecting warning signs of influenza outbreaks earlier. However, supplementation of modern surveillance with search engine information should be approached cautiously. This approach provides valuable insights for digital epidemiology and has the potential for broader application in respiratory infectious disease surveillance. Further research should explore the optimization and customization of search terms for different regions and languages to improve the accuracy of influenza prediction models.

Potential Determinants Contributing to COVID-19 Vaccine Acceptance and Hesitancy in Taiwan: Rapid Qualitative Mixed Methods Study.

Although vaccination has been shown to be one of the most important interventions, COVID-19 vaccine hesitancy remains one of the top 10 global public health challenges worldwide. The objective of this study is to investigate (1) major determinants of vaccine hesitancy, (2) changes in the determinants of vaccine hesitancy at different time periods, and (3) the potential factors affecting vaccine acceptance. This study applied a mixed methods approach to explore the potential determinants contributing to vaccine hesitancy among the Taiwanese population. The quantitative design of this study involved using Google Trends search query data. We chose the search term "" (vaccine), selected "" (Taiwan) as the location, and selected the period between December 18, 2020, and July 31, 2021. The rising keywords related to vaccine acceptance and hesitancy were collected. Based on the responses obtained from the qualitative study and the rising keywords obtained in Google Trends, the 3 most popular keywords related to vaccine hesitancy were identified and used as search queries in Google Trends between December 18, 2020, and July 31, 2021, to generate relative search volumes (RSVs). Lastly, autoregressive integrated moving average modeling was used to forecast the RSVs for the 3 keywords between May 29 and July 31, 2021. The estimated RSVs were compared to the observed RSVs in Google Trends within the same time frame. The 4 prevailing factors responsible for COVID-19 vaccine acceptance and hesitancy were doubts about the government and manufacturers, side effects, deaths associated with vaccination, and efficacy of vaccination. During the vaccine observation period, "political role" was the overarching consideration leading to vaccine hesitancy. During the peak of the pandemic, side effects, death, and vaccine protection were the main factors contributing to vaccine hesitancy. The popularity of the 3 frequently searched keywords "side effects," "vaccine associated deaths," and "vaccine protection" continued to rise throughout the pandemic outbreak. Lastly, the highest Google search queries related to COVID-19 vaccines emerged as "side effects" prior to vaccination, deaths associated with vaccines during the period when single vaccines were available, and "side effects" and "vaccine protection" during the period when multiple vaccines were available. Investigating the key factors influencing COVID-19 vaccine hesitancy appears to be a fundamental task that needs to be undertaken to ensure effective implementation of COVID-19 vaccination. Google Trends may be used as a complementary infoveillance tool by government agencies for future vaccine policy implementation and communication.

Deep Learning Models for Airport Demand Forecasting With Google Trends

Managers gain new insights into how operational benefits can be achieved. Forecasting problems for passenger flow in airports are gaining interest among marketing researchers, but comparison of stochastic optimisation methods via deep learning forecasts with search query data is not yet available in the aviation field. To fill this gap, the current study predicts the demand of Madrid airport demand with Google search query data using H2O deep learning method. The findings indicate that there is a long-term relationship between search queries and actual passenger demand. Besides, search queries “fly to madrid,” and “flights to madrid spain” were found to be the cause of the actual domestic air passenger demand in Madrid. Also, to determine the best forecasting accuracy, stochastic gradient descent (SGD) optimisers were used. Specifically, findings indicate that Adam is a better optimiser increasing forecasting accuracy for Madrid airports.

Analysing and predicting a country’s entrepreneurial activity using insights from entrepreneurs’ digital footprint

PurposeStarting from intention–behaviour models and building upon the growing evidence that aggregated internet search query data represent a good proxy of human interests and intentions. The purpose of this study is to demonstrate that the internet search traffic information related to the selected key terms associated with establishing new businesses, reflects well the dynamics of entrepreneurial activity in a country and can be used for predicting entrepreneurial activity at the national level.Design/methodology/approachTheoretical framework is based on intention–behaviour models and supported by the knowledge spillover theory of entrepreneurship. Monthly data on new business registration from 2018 to 2021 is derived from the open database of the Russian Federal Tax Service. Terms of internet search interest are identified through interviews with the recent founders of new businesses, whereas the internet search query statistics on the identified terms are obtained from Google Trends and Yandex Wordstat.FindingsThe results suggest that aggregated data about web searches related to opening a new business in a country is positively correlated with the dynamics of entrepreneurial activity in the country and, as such, may be useful for predicting the level of that activity.Practical implicationsThe results may serve as a starting point for a new approach to measure, monitor and predict entrepreneurial activities in a country and can help in better addressing policymaking issues related to entrepreneurship.Originality/valueTo the best of the authors’ knowledge, this study is original in its approach and results. Building on intention–behaviour models, this study outlines, to the best of the authors’ knowledge, the first usage of big data for analysing the intention–behaviour relationship in entrepreneurship. This study also contributes to the ongoing debate about the value of big data for entrepreneurship research by proposing and demonstrating the credibility of internet search query data as a novel source of quality data in analysing and predicting a country’s entrepreneurial activity.

Tourism demand forecasting based on user-generated images on OTA platforms

ABSTRACT Tourists’ images of destinations posted on online travel agency (OTA) platforms are important information sources for potential tourists to perceive and construct destination images. These perceptions can then inform travel decisions. We investigated the roles of the aesthetics of user-generated images on OTA platforms in tourism demand forecasting. Specifically, the aesthetics of images of three popular scenic spots in Hong Kong were used to predict tourism demand from the region’s two largest short-haul markets and largest long-haul market using a seasonal autoregressive moving average with exogenous factors (SARIMAX) model. Seasonal naïve, SARIMA, and SARIMAX models involving search query data were taken as benchmarks. Results showed that (1) image aesthetics could help make more accurate tourism demand forecasting; and (2) as an additional variable, image aesthetics could supplement search query-based volume variables to enhance tourism demand forecasting.

Large-Scale Estimation and Analysis of Web Users' Mood from Web Search Query and Mobile Sensor Data.

The ability to estimate the current mood states of web users has considerable potential for realizing user-centric opportune services in pervasive computing. However, it is difficult to determine the data type used for such estimation and collect the ground truth of such mood states. Therefore, we built a model to estimate the mood states from search-query data in an easy-to-collect and non-invasive manner. Then, we built a model to estimate mood states from mobile sensor data as another estimation model and supplemented its output to the ground-truth label of the model estimated from search queries. This novel two-step model building contributed to boosting the performance of estimating the mood states of web users. Our system was also deployed in the commercial stack, and large-scale data analysis with >11 million users was conducted. We proposed a nationwide mood score, which bundles the mood values of users across the country. It shows the daily and weekly rhythm of people's moods and explains the ups and downs of moods during the COVID-19 pandemic, which is inversely synchronized to the number of new COVID-19 cases. It detects big news that simultaneously affects the mood states of many users, even under fine-grained time resolution, such as the order of hours. In addition, we identified a certain class of advertisements that indicated a clear tendency in the mood of the users who clicked such advertisements.

Assessing smoking-related behaviours using massive online search query data

Online search queries have been used in various behaviour studies. As smoking has become a global health issue, studies that assess smoking-related behaviours using online search queries have faced limitations in data utilization and study design. This study conducts comparative analyses to investigate changes in smoking-related behaviours represented by online search volumes. Baidu search queries from 2013 to 2017 were used to examine the search volume containing four groups of smoking-related search keywords. A validation process was used to validate the proposed method by comparing changes in search queries with the observed tobacco consumption. The results show changes in smoking-related behaviours assessed by online search queries at the city level in China. The validation experiments illustrate the consistency between changes in search volumes and tobacco consumption. Thus, online search queries were verified to be an effective instrument for assessing smoking-related behaviours, and this study sheds light on broader behaviour studies and policy assessments.

Web Search Query Volume Correlates With Prescription Volumes of Antidepressants and Antipsychotics in the Netherlands and United Kingdom: An Explorative Study.

The significant increase in Internet availability has resulted in a rise in search queries on health-related topics. Previous research has demonstrated the potential for analyzing web search query volume for nonpsychotropic prescription drugs, while studies on psychotropic drugs remain scarce. The aims of this study were to expand upon this scarce knowledge by investigating the relationship between web search query volumes and prescription volumes of antidepressants and antipsychotics in the United Kingdom and the Netherlands and to gain insight in topics of concern, such as withdrawal symptoms and discontinuation. Data were obtained for the United Kingdom and the Netherlands from January 2010 until January 2021. Prescription volume data for 5 antidepressants (paroxetine, fluoxetine, sertraline, citalopram, venlafaxine) and 5 antipsychotics (quetiapine, olanzapine, clozapine, aripiprazole, and risperidone) were obtained. Web search query volumes and data on related search queries of these substances were acquired from Google Trends. Descriptive statistics and Pearson correlation analyses were performed. A strong, positive, and statistically significant correlation between web search query volume and prescription volume was observed for most included substances in both the Netherlands and the United Kingdom. The search queries related to the included antidepressants and antipsychotics indicate important topics of concern for specific substances, such as withdrawal symptoms and discontinuation. Web search data from Google Trends could potentially be used as a proxy for prescribing trends of antidepressants and antipsychotics and to gain insight in topics of concern of users of these substances. These findings highlight the importance of providing reliable patient information, particularly regarding adverse effects, withdrawal, and discontinuation.

Diverse feature extraction techniques in internet search query to forecast tourism demand: An in-depth comparison

AbstractIt is an effective approach to improve forecasting by extracting effective information from large panels of search query data. Feature extraction techniques (FETs) can extract information from all features by creating new fewer features based on algebraic transformation; however, they have not been extensively investigated and compared for tourism forecasting. We employ five FETs to process multi-dimensional search query data, and build a bunch of models based on econometrics, machine learning, ensemble learning and hybrid methods. The improving performances of FETs based on tourism demand forecasting in Sanya after COVID-19 and in Macau before COVID-19 are evaluated. The results show that forecasting models with FETs outperform the benchmark model SARIMAX without FETs, which demonstrates the efficacy of FETs in search query data extraction. This study provides meaningful guidance for improving the quality of multi-dimensional data and optimizing tourism forecasting.

Impact assessment of immunization and the COVID-19 pandemic on varicella across Europe using digital epidemiology methods: A descriptive study.

Varicella is usually a mild disease in children but may be life-threatening, especially in adolescents and adults. Infection control measures implemented during the Coronavirus Disease 2019 (COVID-19) pandemic may have suppressed varicella transmission, potentially creating an 'immunity debt', particularly in countries without universal varicella vaccination. To assess trends in Google search engine queries for varicella keywords as a proxy for varicella infection rates and to evaluate the effect of universal varicella vaccination on these trends. A further objective was to assess the impact of the COVID-19 pandemic on varicella keyword search query trends in countries with and without universal varicella vaccination. This study used the keyword research tool, Google Trends, to evaluate trends in time series of the relative search query popularity of language-specific varicella keywords in 28 European countries from January 2015 through December 2021. The Google Ads Keyword Planner tool was used to evaluate absolute search volumes from March 2018 through December 2021. The relative search query popularity of varicella keywords displayed marked seasonal variation. In all 28 countries, the relative search query popularity of varicella keywords declined after the start of the COVID-19 pandemic (March 2020), compared with pre-pandemic levels (range, -18% to -70%). From April 2020 to July 2021, a period of intense COVID-19 transmission and infection control, absolute search volumes for varicella keywords were lower than pre-pandemic levels but rebounded after July 2021, when infection control measures were relaxed. This evaluation of search query trends demonstrated that search query data could be used as a proxy for trends in varicella infection rates and revealed that transmission of varicella may have been suppressed during the COVID-19 pandemic. Consideration should be given to using search query data to better understand the burden of varicella, particularly in countries where surveillance systems are inadequate.

Can multi-source heterogeneous data improve the forecasting performance of tourist arrivals amid COVID-19? Mixed-data sampling approach

The coronavirus disease (COVID-19) pandemic has already caused enormous damage to the global economy and various industries worldwide, especially the tourism industry. In the post-pandemic era, accurate tourism demand recovery forecasting is a vital requirement for a thriving tourism industry. Therefore, this study mainly focuses on forecasting tourist arrivals from mainland China to Hong Kong. A new direction in tourism demand recovery forecasting employs multi-source heterogeneous data comprising economy-related variables, search query data, and online news data to motivate the tourism destination forecasting system. The experimental results confirm that incorporating multi-source heterogeneous data can substantially strengthen the forecasting accuracy. Specifically, mixed data sampling (MIDAS) models with different data frequencies outperformed the benchmark models.

A novel hybrid deep learning approachfor tourism demand forecasting

This paper proposes a new hybrid deep learning framework that combines search query data, autoencoders (AE) and stacked long-short term memory (staked LSTM) to enhance the accuracy of tourism demand prediction. We use data from Google Trends as an additional variable with the monthly tourist arrivals to Marrakech, Morocco. The AE is applied as a feature extraction procedure to dimension reduction, to extract valuable information and to mine the nonlinear information incorporated in data. The extracted features are fed into stacked LSTM to predict tourist arrivals. Experiments carried out to analyze performance in forecast results of proposed method compared to individual models, and different principal component analysis (PCA) based and AE based hybrid models. The experimental results show that the proposed framework outperforms other models.

Tracking Population-Level Anxiety Using Search Engine Data: Ecological Study.

Anxiety disorders are the most prevalent mental disorders globally, with a substantial impact on quality of life. The prevalence of anxiety disorders has increased substantially following the COVID-19 pandemic, and it is likely to be further affected by a global economic recession. Understanding anxiety themes and how they change over time and across countries is crucial for preventive and treatment strategies. The aim of this study was to track the trends in anxiety themes between 2004 and 2020 in the 50 most populous countries with high volumes of internet search data. This study extends previous research by using a novel search-based methodology and including a longer time span and more countries at different income levels. We used a crowdsourced questionnaire, alongside Bing search query data and Google Trends search volume data, to identify themes associated with anxiety disorders across 50 countries from 2004 to 2020. We analyzed themes and their mutual interactions and investigated the associations between countries' socioeconomic attributes and anxiety themes using time-series linear models. This study was approved by the Microsoft Research Institutional Review Board. Query volume for anxiety themes was highly stable in countries from 2004 to 2019 (Spearman r=0.89) and moderately correlated with geography (r=0.49 in 2019). Anxiety themes were predominantly long-term and personal, with "having kids," "pregnancy," and "job" the most voluminous themes in most countries and years. In 2020, "COVID-19" became a dominant theme in 27 countries. Countries with a constant volume of anxiety themes over time had lower fragile state indexes (P=.007) and higher individualism (P=.003). An increase in the volume of the most searched anxiety themes was associated with a reduction in the volume of the remaining themes in 13 countries and an increase in 17 countries, and these 30 countries had a lower prevalence of mental disorders (P<.001) than the countries where no correlations were found. Internet search data could be a potential source for predicting the country-level prevalence of anxiety disorders, especially in understudied populations or when an in-person survey is not viable.