Search results

With the rapid growth of the domestic lending industry, assessing whether the borrower of each loan is at risk of default is a pressing issue for financial institutions. Although there are some models that can handle such problems well, there are still some shortcomings in some aspects. The purpose of this paper is to improve the accuracy of credit assessment models.

In this paper, three different stages are used to improve the classification performance of LSTM, so that financial institutions can more accurately identify borrowers at risk of default. The first approach is to use the K-Means-SMOTE algorithm to eliminate the imbalance within the class. In the second step, ResNet is used for feature extraction, and then two-layer LSTM is used for learning to strengthen the ability of neural networks to mine and utilize deep information. Finally, the model performance is improved by using the IDWPSO algorithm for optimization when debugging the neural network.

On two unbalanced datasets (category ratios of 700:1 and 3:1 respectively), the multi-stage improved model was compared with ten other models using accuracy, precision, specificity, recall, G-measure, F-measure and the nonparametric Wilcoxon test. It was demonstrated that the multi-stage improved model showed a more significant advantage in evaluating the imbalanced credit dataset.

In this paper, the parameters of the ResNet-LSTM hybrid neural network, which can fully mine and utilize the deep information, are tuned by an innovative intelligent optimization algorithm to strengthen the classification performance of the model.

The purpose of this research paper is to explore the determinants of patients' service choices between telephone consultation and text consultation in online health communities (OHCs).

This study utilized an empirical model based on the elaboration likelihood model and examined the effect of information, regarding service quality (the central route) and service price (the peripheral route), using online health consultation data from one of the largest OHCs in China.

The logistic regression results indicated that both physician- and patient-generated information can influence the patients' service choices; service price signals will lead patients to cheaper options. However, individual motivations, disease risk and consulting experience change a patients' information processing regarding central and peripheral cues.

Previous researchers have investigated the mechanism of patient behavior in OHCs; however, the researchers have not focused on the patients' choices regarding the multiple health services provided in OHCs. The findings of this study have theoretical and practical implications for future researchers, OHC designers and physicians.

The metaverse enables users to create their own avatars in a shared virtual space, giving rise to a new avatar personality that differs from their real-self personality. The aim of this research is to explore how users' real-self and avatar personalities may affect their behavioral engagement and satisfaction in the metaverse context.

This research applies self-discrepancy theory to investigate how the big five traits of both real-self and avatar personalities influence users' engagement and satisfaction in the metaverse. The present research employed a mixed-methods approach, beginning with a qualitative study to identify prevalent personality cues among users on metaverse social media platforms. Subsequently, a quantitative study was conducted to further validate the findings of the qualitative study.

The results indicated that avatar personality scored higher than the real-self personality in the dimensions of openness, conscientiousness and extraversion, while scored lower in the dimensions of agreeableness and neuroticism. Both real-self and avatar personality traits positively influenced metaverse satisfaction via behavioral engagement in the metaverse. Notably, avatar personality traits had a stronger impact on behavioral engagement compared to real-self personality traits, which further influence metaverse satisfaction.

The present study offers practical insights for metaverse developers and managers to enhance user satisfaction by focusing on users’ big five traits of both real-self and avatar personality. It suggests implementing personalized tools, organizing personality-based social activities and other initiatives to encourage user’s behavioral engagement and ultimately enhance metaverse satisfaction.

Unlike existing research that concentrates on a single facet of personality traits, this research employs a mixed-methods approach to conceptualize users' real-self personality and avatar personality, further exploring their impacts on metaverse satisfaction.

This study explores the effects of virtual streamer characteristics, virtual scene characteristics and streamer image-scene fit on users’ watching intention from the perspective of flow experience.

The survey data for this study were collected from the QQ fan group of virtual streamers between November 26th 2022 and December 5th 2022. The authors survey 274 viewers who have experience of watching virtual streaming and employ the partial least squares structural equation model to test the research hypotheses.

Among the characteristics of virtual streamers, interactivity significantly influences users’ perceived enjoyment and concentration, while vividness only affects perceived enjoyment. In addition, the novelty of the virtual scene has a notable impact on users’ perceived enjoyment and concentration, whereas aesthetic appeal serves as an important indicator solely for concentration. Furthermore, the virtual streamer image-scene fit also affects users’ perceived enjoyment and concentration. Finally, perceived enjoyment and concentration equally contribute to users' watching intention.

This study explores the impact of virtual streamer characteristics, virtual scene characteristics and streamer image-scene fit on users’ watching intention, which enriches the research on user behavioral intention in virtual streaming. Additionally, this study attempts to combine the stimulus-organism-response (S-O-R) model and flow theory in the field of virtual streaming, expanding the research areas. Finally, this study also provides valuable insights for virtual streamers and virtual streaming platforms. By enhancing their virtual personas and optimizing their streaming strategies, virtual streamers can more effectively retain users and maintain audience engagement. Meanwhile, virtual streaming platforms can gain a deeper understanding of user preferences, enabling them to launch high-quality events that sustain user popularity. These efforts collectively contribute to the advancement of the virtual streaming industry.

The purpose of this study is to explore the factors influencing viewers’ engagement intention in travel live streaming (TLS) from a perceived value perspective.

This study used a mixed-methods approach. In Study 1, 48 semistructured interviews were analyzed based on grounded theory and perceived value theory, and a research framework was established to investigate the impact of viewers’ engagement intentions in TLS. In Study 2, partial least squares structural equation modeling (PLS-SEM) was used to empirically validate survey data from 255 TLS viewers.

Through an analysis of the interview content, it was found that the expertise and interaction of the live streamer in TLS as well as the immersion, aesthetics and novelty of the live streaming scene are key influencing factors that affect the engagement of TLS viewers. This finding was confirmed through empirical research.

This research provides practical suggestions for live streamers, TLS platforms and local government to increase viewer engagement. Specifically, it provides methods and directions for the individual improvement of live streamers, further promotes the development and construction of the platform and underscores the importance of government initiatives in policy support and regulatory framework development.

This study focuses on the less-researched field of TLS. Using a mixed-methods approach combining interviews and PLS-SEM, this study explores the key factors that affect the engagement of TLS viewers based on the characteristics of live streamers and live streaming scenes.