Search results

A huge volume of published research articles is available on social media which evolves because of the rapid scientific advances and this paper aims to investigate the research structure of social media.

This study employs an integrated topic modeling and text mining-based approach on 30381 Scopus index titles, abstracts, and keywords published between 2006 and 2021. It combines analytical analysis of top-cited reviews with topic modeling as means of semantic validation. The output sequences of the dynamic model are further analyzed using the statistical techniques that facilitate the extraction of topic clusters, communities, and potential inter-topic research directions.

This paper brings into vision the research structure of social media in terms of topics, temporal topic evolutions, topic trends, emerging, fading, and consistent topics of this domain. It also traces various shifts in topic themes. The hot research topics are the application of the machine or deep learning towards social media in general, alcohol consumption in different regions and its impact, Social engagement and media platforms. Moreover, the consistent topics in both models include food management in disaster, health study of diverse age groups, and emerging topics include drug violence, analysis of social media news for misinformation, and problems of Internet addiction.

This study extends the existing topic modeling-based studies that analyze the social media literature from a specific disciplinary viewpoint. It focuses on semantic validations of topic-modeling output and correlations among the topics and also provides a two-stage cluster analysis of the topics.

This research presents machine learning models for predicting international tourist arrivals in Indonesia during the COVID-19 pandemic using multisource Internet data.

To develop the prediction models, this research utilizes multisource Internet data from TripAdvisor travel forum and Google Trends. Temporal factors, posts and comments, search queries index and previous tourist arrivals records are set as predictors. Four sets of predictors and three distinct data compositions were utilized for training the machine learning models, namely artificial neural networks (ANNs), support vector regression (SVR) and random forest (RF). To evaluate the models, this research uses three accuracy metrics, namely root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE).

Prediction models trained using multisource Internet data predictors have better accuracy than those trained using single-source Internet data or other predictors. In addition, using more training sets that cover the phenomenon of interest, such as COVID-19, will enhance the prediction model's learning process and accuracy. The experiments show that the RF models have better prediction accuracy than the ANN and SVR models.

First, this study pioneers the practice of a multisource Internet data approach in predicting tourist arrivals amid the unprecedented COVID-19 pandemic. Second, the use of multisource Internet data to improve prediction performance is validated with real empirical data. Finally, this is one of the few papers to provide perspectives on the current dynamics of Indonesia's tourism demand.

The purpose of this numerical study is to analyse the character of transition from laminar to chaotic convection in a fluid layer bounded by no‐slip walls in two space dimensions for varying aspect ratio odd‐shaped enclosures consisting of two rectangular chambers, with a linking horizontal enclosure. For a medium Prandtl number fluid (Pr=7), the numerical solution of two‐dimensional Navier‐Stokes momentum and energy equations with Bousinessq approximation has been carried out. It has been found that there are finite Rayleigh numbers Ra₁, Ra₂ and Ra₃ for the onset of single, two and multiple frequency oscillatory motion at different spatial locations in the enclosure. As Ra is further increased period doubling is observed. The onset of strong chaos appears when Ra=Ra₃. This system does not revert to steady state convection at high Ra as observed by other researchers for the case of Rayleigh‐Benard convection. Moreover, the period doubling transition process is consistent with the scenario of Ruelle, Takens and Newhouse. As Ra increases, the power spectrum, and time series of various dynamical variable signals, etc. all show an increasing degree of characteristics of chaos.

Although the business model field of study has been a focus of attention for both researchers and practitioners within the past two decades, it still suffers from concern about its identity. Accordingly, this paper aims to clarify the intellectual structure of business model through identifying the research clusters and their sub-clusters, the prominent relations and the dominant research trends.

This paper uses some common text mining methods including co-word analysis, burst analysis, timeline analysis and topic modeling to analyze and mine the title, abstract and keywords of 14,081 research documents related to the domain of business model.

The results revealed that the business model field of study consists of three main research areas including electronic business model, business model innovation and sustainable business model, each of which has some sub-areas and has been more evident in some particular industries. Additionally, from the time perspective, research issues in the domain of sustainable development are considered as the hot and emerging topics in this field. In addition, the results confirmed that information technology has been one of the most important drivers, influencing the appearance of different study topics in the various periods.

The contribution of this study is to quantitatively uncover the dominant knowledge structure and prominent research trends in the business model field of study, considering a broad range of scholarly publications and using some promising and reliable text mining techniques.

Using local historical data, the purpose of this paper is to compile a chronology of high‐magnitude snow avalanches to illustrate the effectiveness of information published in newspapers in assisting the management of natural hazards.

Major transportation lines parallel the southern boundary of Glacier National Park, Montana, USA, an area where snow avalanches occur frequently and occasionally block transportation corridors. A 1986 study presented an avalanche chronology for the study area based on information collected from the 1946 to 1982 issues of a local weekly newspaper. We extend that existing data set here by using the same newspaper, recording avalanche occurrences and impacts reported in the newspaper from 1982 to 2005.

The newly updated chronology is presented, with trends and temporal patterns for the entire 1946‐2005 data set analyzed. A decrease in reported avalanches, from the 1960s onward, is noted. Additionally, reported avalanches shifted from occurring most frequently in February to January in the last 20 years.

The results of this research illustrate the use of newspaper reports as an inexpensive, but effective, way to compile a chronology of high‐magnitude snow avalanches. This research method tends to underreport smaller magnitude events that do not affect the transportation linkages.

Snow avalanche managers could use this method to compile a chronology of events when other, more traditional techniques, are not available or too expensive.

This paper uses a rarely utilized but inexpensive and widely available data source to construct a 59‐year avalanche chronology in an area constantly threatened by snow avalanches.

The purpose of this paper is to explore the sensitivity of economic efficiency rankings of water businesses to the choice of alternative physical and accounting capital input measures.

Data envelopment analysis (DEA) was used to compute efficiency rankings for government-owned water businesses from the state of Victoria, Australia, over the period 2005/2006 through 2012/2013. Differences between DEA models when capital inputs were measured using either: statutory accounting values (historic cost and fair value), physical measures, or regulatory accounting values, were scrutinised.

Depending on the choice of capital input, significant variation in efficiency scores and the ranking of the top (worst) performing firms was observed.

Future research may explore the generalisability of findings to a wider sample of water utilities globally. Future work can also consider the most reliable treatment of capital inputs in efficiency analysis.

Regulators should be cautious when using economic efficiency data in benchmarking exercises. A consistent approach to account for the capital stock is needed in the determination of price caps and designing incentives for poor performers.

DEA has been widely used to explore the role of ownership structure, firm size and regulation on water utility efficiency. This is the first study of its kind to explore the sensitivity of DEA to alternative physical and accounting capital input measures. This research also improves the conventional performance measurement in water utilities by using a bootstrap procedure to address the deterministic nature of the DEA approach.

Although acknowledged as a principal dimension in the context of text mining, time has yet to be formally incorporated into the process of visually representing the relationships between keywords in a knowledge domain. This paper aims to develop and validate the feasibility of adding temporal knowledge to a concept map via pair-wise temporal analysis (PTA).

The paper presents a temporal trend detection algorithm – vector space model – designed to use objective quantitative pair-wise temporal operators to automatically detect co-occurring hot concepts. This PTA approach is demonstrated and validated without loss of generality for a spectrum of information technologies.

The rigorous validation study shows that the resulting temporal assessments are highly correlated with subjective assessments of experts (n = 136), exhibiting substantial reliability-of-agreement measures and average predictive validity above 85 per cent.

Using massive amounts of textual documents available on the Web to first generate a concept map and then add temporal knowledge, the contribution of this work is emphasized and magnified against the current growing attention to big data analytics.

This paper proposes a novel knowledge discovery method to improve a text-based concept map (i.e. semantic graph) via detection and representation of temporal relationships. The originality and value of the proposed method is highlighted in comparison to other knowledge discovery methods.

The purpose of this paper is to identify salient topic categories and outline their evolution patterns and temporal trends in microblogs on a public health emergency across different stages. Comparisons were also examined to reveal the similarities and differences between those patterns and trends on microblog platforms of different languages and from different nations.

A total of 459,266 microblog entries about the Ebola outbreak in West Africa in 2014 on Twitter and Weibo were collected for nine months after the inception of the outbreak. Topics were detected by the latent Dirichlet allocation model and classified into several categories. The daily tweets were analyzed with the self-organizing map technique and labeled with the most salient topics. The investigated time span was divided into three stages, and the most salient topic categories were identified for each stage.

In total, 14 salient topic categories were identified in microblogs about the Ebola outbreak and were summarized as increasing, decreasing, fluctuating or ephemeral types. The topical evolution patterns of microblogs and temporal trends for topic categories vary on different microblog platforms. Twitter users were keen on the dynamics of the Ebola outbreak, such as status description, secondary events and so forth, while Weibo users focused on background knowledge of Ebola and precautions.

This study revealed evolution patterns and temporal trends of microblog topics on a public health emergency. The findings can help administrators of public health emergencies and microblog communities work together to better satisfy information needs and physical demands by the public when public health emergencies are in progress.

This paper aims to identify what drives the temporal reduction in the value relevance of earnings documented in the literature. Is it the increasing noise in stock returns over time, noise in earnings, or both?

The authors develop hypotheses from the lead/lag structure between stock returns and accounting earnings and perform empirical tests using data from annual COMPUSTAT and monthly CRSP over the sample period of 39 years (1970‐2008).

The test results show that increasing noise in stock returns over time is primarily responsible for the temporal reduction of R² in regressions of returns on earnings. Additional analysis shows weak evidence that both the noise in returns and the noise in earnings are responsible for the declining association between earnings and returns in a sub‐period (1970‐1982).

The R²‐based methodology has limitations because, as Gu points out, regression R²s might be incomparable across samples. The findings suggest that future research should control for the effects of the temporal increase in market noise before making value relevance inferences from the declining association between earnings and returns.

The paper contributes to the limited body of research on noise in stock returns as the main driver for the temporal reduction in value relevance of earnings.

This paper brings focus and attention to the aspect of time within health information behaviour. The purpose of this paper is to critically assess and present strengths and weaknesses of utilising the infodemiology approach and metrics as a novel way to examine temporal variations and patterns of online health information behaviour. The approach is shortly exemplified by presenting empirical evidence for temporal patterns of health information behaviour on different time-scales.

A short review of online health information behaviour is presented and methodological barriers to studying the temporal nature of this behaviour are emphasised. To exemplify how the infodemiology approach and metrics can be utilised to examine temporal patterns, and to test the hypothesis of existing rhythmicity of health information behaviour, a brief analysis of longitudinal data from a large discussion forum is analysed.

Clear evidence of robust temporal patterns and variations of online health information behaviour are shown. The paper highlights that focussing on time and the question of when people engage in health information behaviour can have significant consequences.

Studying temporal patterns and trends for health information behaviour can help in creating optimal interventions and health promotion campaigns at optimal times. This can be highly beneficial for positive health outcomes.

A new methodological approach to study online health information behaviour from a temporal perspective, a phenomenon that has previously been neglected, is presented. Providing evidence for rhythmicity can complement existing epidemiological data for a more holistic picture of health and diseases, and their behavioural aspects.