Search results

With the development of intelligent technology, deep learning has made significant progress and has been widely used in various fields. Deep learning is data-driven, and its training process requires a large amount of data to improve model performance. However, labeled data is expensive and not readily available.

To address the above problem, researchers have integrated semi-supervised and deep learning, using a limited number of labeled data and many unlabeled data to train models. In this paper, Generative Adversarial Networks (GANs) are analyzed as an entry point. Firstly, we discuss the current research on GANs in image super-resolution applications, including supervised, unsupervised, and semi-supervised learning approaches. Secondly, based on semi-supervised learning, different optimization methods are introduced as an example of image classification. Eventually, experimental comparisons and analyses of existing semi-supervised optimization methods based on GANs will be performed.

Following the analysis of the selected studies, we summarize the problems that existed during the research process and propose future research directions.

This paper reviews and analyzes research on generative adversarial networks for image super-resolution and classification from various learning approaches. The comparative analysis of experimental results on current semi-supervised GAN optimizations is performed to provide a reference for further research.

Over energy consumption is one of the causes of global warming and climate change. To deal with this issue, using energy-efficient appliances is strongly encouraged and cultivating consumer loyalty toward energy-efficient appliances is crucial for long-term sustainability. This study investigates the effects of multiple dimensions of perceived value on consumer satisfaction and three outcomes of consumer loyalty (i.e. willingness to pay a premium, repurchase intention and word-of-mouth intention), considering the moderating role of the product category.

Quota sampling based on age and gender and snowball sampling methods were applied to recruit 423 participants for this study. A combination of partial least squares structural equation modeling (PLS-SEM), importance-performance map analysis (IPMA) and necessary condition analysis (NCA) was employed to examine the proposed model.

Functional, price, emotional and environmental values positively impact consumer satisfaction, enhancing the three dimensions of consumer loyalty. The product category moderates the effects of perceived value dimensions on consumer satisfaction. Moreover, the IPMA results highlight that functional value and environmental value are the most essential but underperforming value attributes. The NCA results indicate that social value is a necessary condition for consumer satisfaction.

This study is one of the pioneers in integrating PLS-SEM, IPMA and NCA approaches to comprehensively unpack the relationships between perceived value dimensions, consumer satisfaction and consumer loyalty in the context of energy-efficient appliances. The findings offer theoretical and practical importance for academics, retailers, producers and policymakers to encourage consumer loyalty toward energy-efficient appliances.

The purpose of this study was to investigate the impact of artificial intelligence (AI) adoption on knowledge workers' innovative work behaviors (IWB), as well as the mediating role of stress appraisal and the moderating role of individual learning abilities.

This study analyzed the questionnaire results of 313 knowledge workers, and data analysis was conducted by using SPSS 25.0, SPSS 25.0 macro-PROCESS and AMOS 28.0.

This study found that AI adoption has a double-edged sword effect on knowledge workers' IWB. Specifically, AI adoption can promote IWB by enhancing knowledge workers' challenging stress appraisal, while inhibiting IWB by fostering their hindering stress appraisal. Moreover, individual learning ability significantly moderated the relationship between AI adoption and stress appraisal, which further influenced IWB.

This study integrates the conflicting findings of previous studies and proposes a comprehensive theoretical model based on the theory of cognitive appraisal of stress. This study enriches the research on AI in the field of knowledge management, especially extending the understanding of the relationship between AI adoption and knowledge workers’ IWB by unraveling the psychological mechanisms and behavior outcomes of users' technology usage. Additionally, we provide new insights and suggestions for organizations to seek the cooperation and support of employees in introducing new technologies or driving intelligent transformation.

This paper aims to provide a broad overview of the corpus of studies on digital finance in relation to small and medium enterprises (SMEs) and startups.

Bibliometric analysis was used, allowing to investigate the relevant literature (735 articles). In accordance with best practices, relevant articles were identified on the topic following the PRISMA 2020 framework that ensures reproducible and rigorous results. The search then proceeds with performance analysis, identifying key trends at the intersection of research fields, including distribution of articles by year, citations by year, most cited contributions and most cited and prolific authors. This is followed by analyses of co-citation, co-authorship and co-occurrence with a detailed description of the thematic clusters identified.

Performance analysis shows that scholarly output covers a 12-year period, starting in 2011, and demonstrates a growing interest in this topic. Co-occurrence analysis reveals a significant intellectual structure which allows numerous knowledge gaps to emerge, and these offer new opportunities to be addressed in future research.

This study uniquely focuses on the evolution of the research domain related to digital finance associated with SMEs and startups. It provides implications for practitioners and avenues that researchers can develop in the future to produce impactful studies.

The purpose of this study is to investigate the influence of attitude, subjective norms, perceived behavioral control and perceived benefits on consumers’ intention to purchase energy-efficient appliances (EEAs) in an emerging market.

A total of 400 samples were collected via a self-administered questionnaire distributed in Kuala Lumpur, Malaysia. The data was analyzed using multiple regression analysis to assess the hypothesized relationships.

The results of this study reveal that attitude is the key predictor of consumers’ intention to purchase EEAs, followed by perceived benefits of EEAs. This positive attitude stems from the realization that reducing electricity consumption is not only crucial but also a commendable and valuable practice. They can contribute to the collective effort to mitigate climate change, reduce carbon emissions and conserve valuable natural resources. Their positive attitude toward EEAs reflects their sense of responsibility, mindfulness and desire to make an important contribution to promoting sustainability and creating a better future for generations to come.

Energy-efficient companies, retailers and marketers should implement a range of appealing cash rebate programs to stimulate immediate sales, foster future purchases of EEAs and reshape the perception that EEAs are costly. By implementing such rebate plans, the perceived financial burden on consumers is alleviated, resulting in improved attitudes toward EEAs and heightened recognition of their perceived benefits. Consequently, this encourages a surge in demand for EEAs, thereby further propelling the growth of the industry. These sustainable practices align with SDG 9 (Industry, Innovation and Infrastructure) and SDG 12 (Responsible Consumption and Production).

This study stands out for its exceptional contribution to theory, as it applies the theory of planned behavior as the underpinning theory and simultaneously integrates the perceived benefits of EEAs into the proposed model, aiming to foster consumers’ intention to purchase EEAs. What sets this study apart is its examination of an emerging market, which complements and expands upon previous research predominantly conducted in developed (Western) economies.

Research into older adults' use of social media remains limited. Driven by increasing digitalisation in China, the authors focus on Chinese older adults (aged 60–75)’ use of WeChat.

This study used a qualitative interpretive approach and interviewed Chinese older adults to uncover their social practices of WeChat use in everyday life.

By using social practice theory (SPT), the paper unfolds Chinese older adults' social practices of WeChat use in everyday life and reveals how they adopt and resist the drastic changes in Chinese society.

The study contributes to new understandings of SPT from technology use by emphasising the dynamic characteristics of its three elements. The authors synthesise both adoptions and resistance in SPT and highlight the importance of understanding three elements interdependently within specific contexts, which are conditioned by structure and agency.

User discontinuance on short-video platform has become increasingly prevalent in recent years. Short-video discontinuance refers to reduced use, controlled use or suspended use of the short-video platform. In this study, we examined factors associated with discontinuance behavior on short-video platform.

From the perspective of stressor–strain–outcome (SSO), we put forward a theoretical model integrating perceived information overload and perceived system feature overload (stressors), dissatisfaction (psychological strain), flow experience and regret to explain discontinuance behavior on short-video platform (behavioral outcome). We collected 482 survey data from Douyin users in China, and empirically examined the proposed research model via Partial least squares structural equation modeling (PLS-SEM) technique.

Our results demonstrated that perceived system feature overload exerts a positive effect on perceived information overload. Perceived system feature overload has a stronger influence on dissatisfaction than perceived information overload. Regret increases user dissatisfaction, while flow experience decreases user dissatisfaction. We also discovered that dissatisfaction and regret have significant positive effects on discontinuance behavior. Interestingly, flow exerts no significant influence on discontinuance behavior.

This study enriches the body of knowledge on social media discontinuance by revealing the interaction and effects of flow experience, dissatisfaction and regret on discontinuance. This study also extends the understanding on the complex role of flow experience in leading to social media discontinuance. Additionally, this study deepens the research on the interaction between perceived system feature overload and perceived information overload as well as their different influences on negative emotion.

The aim of this study is to examine the ownership impact on firm sustainable investments (FSIs). In particular, this research examines the link between institutional investor ownership (IIO), managerial ownership (MOWN) and FSIs in the tourism industry in Malaysia.

This study uses a data set of 346 firm-year observations from 2008 to 2020 and applies feasible generalized least squares (FGLS) regression analysis. The study sample is based on tourism firms listed on Bursa Malaysia (the Malaysian Stock Exchange).

There is a significant positive association between IIO and FSIs. When IIO is classified into foreign (FIIO) and local (LIIO), this significant association is mainly driven by FIIO. In addition, there is a significant, positive association between managerial ownership (MOWN) and firm sustainable investments (FSIs). These findings imply that firm ownership has an influence on FSIs in the tourism industry.

This is the first attempt to consider IIO and MOWN simultaneously in a single model estimation. The findings contribute to emerging capital markets where the involvement of ownership concentration in the governance of publicly listed firms is a common practice.

This paper examines how firms respond to local government’s environment initiatives through textual analysis of government work reports (GWRs). This study aims to provide insights into how firms strategically respond to government’s environmental initiatives through their disclosure and investment practices.

This study uses a textual analysis of GWRs from China’s provinces. The frequency and change rate of environmental keywords in these reports are used as a measure of the government’s environmental initiatives.

This study finds that environmental disclosure scores in environmental, social and governance (ESG) reports increase with the frequency or change rate of environmental keywords in provincial GWRs. This effect is more pronounced for non-state-owned enterprises, firms in highly marketized provinces or those listed in a single capital market. However, there is no significant relationship between firms’ environmental investments and government initiatives, except for cross-listed firms in provinces with consistently high frequency of environmental keywords in their GWRs.

The findings indicate that government environmental initiatives can shape firms’ disclosure behaviors, yet have limited influence on investment decisions, suggesting that environmental disclosure could potentially be opportunistic. This underscores the need for more effective strategies to stimulate firms’ environmental investments.

This study provides valuable insights into the differential impacts of government environmental initiatives on firms’ disclosure and investment behaviors, contributing to the understanding of corporate environmental responsibility in the context of government initiatives.

Corporate ESG performance has attracted widespread attention from various sectors of society. This paper aims to investigate whether analysts’ ESG attention can convey additional information to the market and consequently influence stock pricing efficiency.

Using A-share listed companies from 2014 to 2021 as the research subjects, this paper employs a deep learning algorithm, word2vec, to construct an ESG dictionary. Text analysis is then applied to create an analysts’ ESG attention index, delving into its impact on stock pricing efficiency.

Empirical research reveals that: (1) Analysts' ESG attention effectively enhances stock pricing efficiency, with a more significant impact from analysts’ attention to environmental (E) and social (S) factors compared to governance (G); (2) Further analysis indicates that this effect becomes more pronounced when there is higher disparity in corporate ESG ratings, greater marketization in the province where the company is located, and a higher institutional ownership percentage and (3) The mechanism by which analysts' ESG attention influences stock pricing efficiency is through an elevation in investor attention and stock liquidity. Additionally, it is observed that analysts prioritize ESG information to enhance their reputation and business capabilities.

From the perspective of ESG rating divergence, this paper innovatively uses analyst reports to construct ESG attention indicators and analyzes their impact on the efficiency of stock pricing.