Search results

This study aims to construct a sentiment series generation method for danmu comments based on deep learning, and explore the features of sentiment series after clustering.

This study consisted of two main parts: danmu comment sentiment series generation and clustering. In the first part, the authors proposed a sentiment classification model based on BERT fine-tuning to quantify danmu comment sentiment polarity. To smooth the sentiment series, they used methods, such as comprehensive weights. In the second part, the shaped-based distance (SBD)-K-shape method was used to cluster the actual collected data.

The filtered sentiment series or curves of the microfilms on the Bilibili website could be divided into four major categories. There is an apparently stable time interval for the first three types of sentiment curves, while the fourth type of sentiment curve shows a clear trend of fluctuation in general. In addition, it was found that “disputed points” or “highlights” are likely to appear at the beginning and the climax of films, resulting in significant changes in the sentiment curves. The clustering results show a significant difference in user participation, with the second type prevailing over others.

Their sentiment classification model based on BERT fine-tuning outperformed the traditional sentiment lexicon method, which provides a reference for using deep learning as well as transfer learning for danmu comment sentiment analysis. The BERT fine-tuning–SBD-K-shape algorithm can weaken the effect of non-regular noise and temporal phase shift of danmu text.

Today’s one of the most important difficulties is tackling climate change and its effects on the environment. The Paris Agreement states that nations must balance the amount of greenhouse gases they emit and absorb until 2050 to contribute to the mitigation of greenhouse gases and to support sustainable development. According to the agreement, each country must determine, plan and regularly report on its contributions. Thus, it is important for the countries to predict and analyze their net zero performances in 2050. Therefore, the aim of this study is to evaluate European Continent Countries' net zero performances at the targeted year.

The European Continent Countries that ratified the Paris Agreement are specified as decision making units (DMUs). Input and output indicators are specified as primary energy consumption, freshwater withdrawals, gross domestic product (GDP), carbon-dioxide (CO₂) and nitrous-oxide (N₂O) emissions. Data from 1980 to 2019 are obtained and forecasted using autoregressive integrated moving average (ARIMA) until 2050. Then, the countries are clustered based on the forecasts of primary energy consumption and freshwater withdrawals using k-means algorithm. As desirable and undesirable outputs arise simultaneously, the performances are computed using Pure Environmental Index (PEI) and Mixed Environmental Index (MEI) data envelopment analysis (DEA) models.

It is expected that by 2050, CO₂ emissions of seven countries remain constant, N₂O emissions of seven countries remain stable and five countries’ both CO₂ and N₂O emissions remain constant. While it can be seen as success that many countries are expected to at least stabilize one emission, the likelihood of achieving net zero targets diminishes unless countries undertake significant reductions in emissions. According to the results, in Cluster 1, Turkey ranks last, while France, Germany, Italy and Spain are efficient countries. In Cluster 2, the United Kingdom ranks at last, while Greece, Luxembourg, Malta and Sweden are efficient countries.

In the literature, generally, CO₂ emission is considered as greenhouse gas. Moreover, none of the studies measured the net-zero performance of the countries in 2050 employing analytical techniques. This study objects to investigate how well European Continent Countries can comply with the necessities of the Agreement. Besides CO₂ emission, N₂O emission is also considered and the data of European Continent Countries in 2050 are estimated using ARIMA. Then, countries are clustered using k-means algorithm. DEA models are employed to measure the performances of the countries. Finally, forecasts and models validations are performed and comprehensive analysis of the results is conducted.

To plan the future university campus, campus executives need decision-making support from theory and practice. Matching the static campus (supply) with the dynamic (demand) - while safeguarding spatial quality and sustainability - requires management information from similar organizations. This study presents an evidence-based briefing approach to support decision-makers of individual universities with management information when making decisions for their future campus.

For the proposed evidence-based briefing approach, the continuous Designing an Accommodation Strategy (DAS) framework is used in a mixed-method research design to evaluate the past to plan for the future. Five campus themes and three campus models (solid, liquid, and gas) are introduced to describe the development and diversification of university campuses and their impact across different university building types. Based on this theoretical framework, first, qualitative interview data are analyzed to understand which standards campus managers expect; second, a quantitative project database is used to demonstrate what is actually realized.

The findings demonstrate that remote working and online education will become more common. Academic workplaces and learning environments are more adaptive to changes than laboratory spaces. The analyses reveal different effective space use strategies to meet the current demand: they include space-efficient mixed-use buildings, and mono-functional generic educational and office spaces. These results show that operationalized evidence-based briefing can help design the future campus.

The study adds knowledge during a critical (post-COVID) period when decision-makers need evidence from others to adapt their campus management strategies to hybrid and sustainable ambitions.

The purpose of this paper is to propose a scenario-based grey methodology using clustering and optimizing with imprecise and uncertain body size data in an emergency assembly point area to assign the people on a campus to reach the emergency assembly points under uncertain disaster times.

Grey clustering and a new grey p-median linear programming model are developed to determine which units to assign to the pre-determined assembly points for a main campus in case of a disaster. The models have two scenarios: 70 and 100% occurrence capacities of administrative and academic personnel and students.

In this study, the academic and administrative units have been assigned to determine five different emergency assembly points on the main campus by using the numbers of the academic and administrative personnel and student and distances of the units to the assembly point areas of each other. The alternative solutions are obtained effectively by evaluating capacity utilization rates in the scenarios.

It is often unclear when disasters can occur and therefore, a preliminary preparation time must be required to minimize the risk. In the case of natural, man-made (unnatural) or technological disasters, the people are required to defend themselves and move away from the disaster area as soon as possible in a proper direction. The proposed assignment model yields a final solution that effectively eliminates uncertainty regarding the selection of emergency assembly points for administrative and academic staff as well as students, in the event of disasters.

Grey clustering suggests an assignment plan and concurrently, an investigation is underway utilizing the grey p-median linear programming model. This investigation aims to optimize various scenarios and body sizes concerning emergency assembly areas. All campus users who are present at the disaster in units of the campus are getting uncertainty about which emergency assembly point to use, and with this study, the vital risks aim to be ultimately reduced with reasonable plans.

The purpose of this paper is to guide commercial and policy action to improve smallholder Indonesian cattle systems.

A survey (n = 304) of smallholder cattle farms in six villages in two districts of Nusa Tenggara Barat, Indonesia. Principal component analysis (PCA) and cluster analysis (CA) were employed to classify cattle farms into business models according to observed innovation. Differences between business models were identified using a one-way-analysis-of-variance (ANOVA).

Four business models were identified, representing profiles of innovation adoption and elements of business models, socio-economic characteristics, farming system and performance variables including revenue, cost and profit. The business models display a range of orientation to buyer requirements and a range of approaches to production, indicating a need to promote in a variety of ways the change from supply-push to demand-pull in the cattle value chain.

This study offers guidance on how business models might be strengthened over time, by using simple indicators of performance and the models' linkage to innovation in the context of each business model. The business models developed here, and refinements to them based on localised conditions, offer a targeted and accelerated pathway to improved performance in smallholder systems.

This study proposed a novel approach to the recognition of business models based on innovation.

COVID-19 continues to spread, and cause increasing deaths. Physicians diagnose COVID-19 using not only real-time polymerase chain reaction but also the computed tomography (CT) and chest x-ray (CXR) modalities, depending on the stage of infection. However, with so many patients and so few doctors, it has become difficult to keep abreast of the disease. Deep learning models have been developed in order to assist in this respect, and vision transformers are currently state-of-the-art methods, but most techniques currently focus only on one modality (CXR).

This work aims to leverage the benefits of both CT and CXR to improve COVID-19 diagnosis. This paper studies the differences between using convolutional MobileNetV2, ViT DeiT and Swin Transformer models when training from scratch and pretraining on the MedNIST medical dataset rather than the ImageNet dataset of natural images. The comparison is made by reporting six performance metrics, the Scott–Knott Effect Size Difference, Wilcoxon statistical test and the Borda Count method. We also use the Grad-CAM algorithm to study the model's interpretability. Finally, the model's robustness is tested by evaluating it on Gaussian noised images.

Although pretrained MobileNetV2 was the best model in terms of performance, the best model in terms of performance, interpretability, and robustness to noise is the trained from scratch Swin Transformer using the CXR (accuracy = 93.21 per cent) and CT (accuracy = 94.14 per cent) modalities.

Models compared are pretrained on MedNIST and leverage both the CT and CXR modalities.

This paper presents a conceptual framework based on an extensive literature review. The aim of this study is to deepen understanding of the relationship between carbon performance and the financial market by applying qualitative research approaches.

The investigation has identified 372 articles sourced from Scopus databases, subjecting the bibliographic data to a comprehensive qualitative–quantitative analysis. The research uses established protocols for a structured literature review, adhering to PRISMA guidelines, machine learning-based structural topic modelling using Python and bibliometric citation analysis.

The results identified the leading academic authors, institutions and countries concerning carbon performance and financial markets literature. Quantitative studies dominate this research theme. The study has identified six knowledge clusters using topic modelling related to environmental reporting; price drivers of carbon markets; environmental policy and capital markets; financial development and carbon emissions; carbon risk and financial markets; and environmental performance and firm value. The results of the study also present the opportunities associated with carbon performance and the financial market and propose future research agendas on research through theory, characteristics, context and methodology.

The results of the study offer insights to practitioners, researchers and academicians regarding scientific development, intricate relationships and the complexities involved in the intersection of carbon performance and financial markets. For policymakers, a better understanding of carbon performance and financial markets will contribute to designing policies to set up priorities for countering carbon emissions.

The study highlights the critical areas that require attention to limit greenhouse gas emissions and promote decarbonisation effectively. Policymakers can leverage these insights to develop targeted and evidence-based policies that facilitate the transition to a more sustainable and low-carbon economy.

The study initially attempts to discuss the research stream on carbon performance and financial markets literature from a systematic literature review.

The purpose of this research is to identify the most important attributes of metaverse influencers and examine their impact on customer engagement and social glue.

Three studies (one qualitative and two quantitative) were conducted to understand the phenomenon better. The qualitative study (Study 1) was conducted to identify the antecedents of the theoretical model, which was tested in Study 2 using the covariance-based structural equation modelling (CB-SEM) technique. Study 3 then divided the respondents based on the metaverse influencer attribute preferences.

Results of Study 1 revealed the six most influential attributes of metaverse influencers: physical attractiveness, social attractiveness, perceived credibility, metaverse-influencer fit, intimacy and attitude homophily. Further, Study 2 validated that attractiveness and perceived credibility enhance engagement. Also, the results revealed that intimacy, perceived credibility and homophily enhance social glue. Moreover, parasocial relationships mediate the association between intimacy, attitude homophily, perceived credibility and (engagement and social glue). The conditional indirect effect of physical attractiveness, social attractiveness and metaverse–influencer fit on (engagement and social glue) via parasocial relationships at different high and low levels of self-discrepancy was significant. Finally, Study 3 used latent class analysis to reveal different clusters of metaverse users.

This research enriches our understanding of metaverse influencers, contributing to the influencer marketing literature. It offers actionable insights for marketers by elucidating key influencer attributes, aiding in enhancing engagement and social glue.

This study aims to examine the characteristics of cross-border central bank digital currencies (CBDCs) while pinpointing research trends and adoption variables at both individual and macroeconomic levels. Additionally, it delves into the impact of terminology within CBDC-related scholarly literature themes.

The authors perform a bibliometric study using the metadata of academic papers about CBDC from ScienceDirect, Scopus and Web of Science (WoS), three reputable research databases. Word maps are produced using VOSviewer, an open-source bibliometric analytics program, to find pertinent and predominate words and phrases based on their frequency, placement, connection and co-occurrence. Additionally, the authors use the R programing language to assess the Jaccard similarity between bibliometric metadata and the financial terms in the Loughran-McDonald Master Dictionary (LMMD).

The study pinpoints the factors that affect CBDC adoption at the micro and macroeconomic levels. Insights into prospective future study themes are provided by the analysis of the metadata corpus, which shows significant and predominate words/phrases and themes in CBDC literature. Notably, the relatively low Jaccard similarity scores in the scholarly literature on CBDC-related topics across all three bibliometric databases suggest a restricted concentration on financial issues. This shows that CBDC research is still in its early stages and that there are still many undiscovered financial aspects.

The identification of literature’s themes using dominant and pertinent words based on bibliometric metadata, considering factors such as frequency and co-occurrence, enriches the evolving field of meta-analysis. Additionally, the use of the Jaccard index to assess the coverage of financial terms within bibliometric metadata represents a unique approach, shedding light on the distinctive aspects of CBDC research.

The goal of this article is to provide an overview of healthcare entrepreneurship, both in terms of its current trends and future directions.

The article engages in a systematic review of extant research on healthcare entrepreneurship using the scientific procedures and rationales for systematic literature reviews (SPAR-4-SLR) as the review protocol and bibliometrics or scientometrics analysis as the review method.

Healthcare entrepreneurship research has fared reasonably well in terms of publication productivity and impact, with diverse contributions coming from authors, institutions and countries, as well as a range of monetary and non-monetary support from funders and journals. The (eight) major themes of healthcare entrepreneurship research revolve around innovation and leadership, disruption and technology, entrepreneurship models, education and empowerment, systems and services, orientations and opportunities, choices and freedom and policy and impact.

The article establishes healthcare entrepreneurship as a promising field of academic research and professional practice that leverages the power of entrepreneurship to advance the state of healthcare.

The article offers a seminal state of the art of healthcare entrepreneurship research.