Purpose: Live streaming commerce has emerged as an essential strategy for vendors to effectively promote their products due to its unique content presentation and real-time interaction. This study aims to investigate the influence of viewer-streamer interaction and viewer-viewer interaction on consumer trust and the subsequent impact of trust on consumers' purchase intention within the live streaming commerce context.Design/methodology/approach: A survey questionnaire was conducted to collect data, and 403 experienced live streaming users in China were recruited. Covariance-based structural equation modeling (CB-SEM) was used for data analysis.Findings: The results indicated that viewer-streamer interaction factors (i.e., personalization and responsiveness) and viewer-viewer interaction factors (i.e., co-viewer involvement and bullet-screen mutuality) significantly influence trust in streamers and co-viewers. Additionally, drawing on trust transfer theory, trust in streamers and co-viewers positively influences trust in products, while trust in co-viewers also positively influences both trust in streamers and products. Furthermore, all three forms of trust positively impact consumers' purchase intentions.Originality/value: This study enriches the extant literature by investigating interaction-based trust-building mechanisms and uncovering the transfer relationships among three trust targets (streamers, co-viewers and products). Furthermore, this study provides some practical guidelines to the streamers and practitioners for promoting consumers’ trust and purchase intention in live streaming commerce.

Virtual streamers, as a typical application of AI-enabled digital humans, are increasingly being utilized in live-streaming commerce due to technological advancements and industry innovations. Although virtual streamers present several benefits, there is potential for adverse effects when they do not align with consumer expectations. Drawing upon expectancy violations theory, this study developed a theoretical model to explore whether and how consumers’ expectation violations during human–virtual streamer interactions affect consumers’ discontinuance behavior. Through an online questionnaire survey of 307 Chinese consumers with prior experience interacting with virtual streamers, this study used a partial least squares structural equation model to analyze the research model. The empirical results indicated that professionalism expectation violation, empathy expectation violation, and responsiveness expectation violation positively influenced consumers’ distrust and dissatisfaction, which subsequently led to discontinuance behavior. This study contributes to the literature on live-streaming commerce, human–AI interaction, and expectancy violation theory. Furthermore, the findings offer valuable insights for practitioners in the field of live-streaming commerce by enabling them to formulate preventive or remedial strategies to mitigate potential negative outcomes when implementing virtual streamers.

With the explosive growth of data, personalized recommendation systems have become an essential component in digitally-driven society. However, one of the main challenges faced by traditional recommendation algorithms is the difficulty in obtaining accurate user preferences, especially for users with limited historical interaction data., which ultimately impacts the performance of these methods. To address this challenge, the paper proposes a recommendation method, named STRGAN (Social trust relationships Generative Adversarial Network), it leverages the advantages of Generative Adversarial Networks (GANs) to tackle the data sparsity problem, by integrating user ratings and social relationships. By incorporating both types of information, STRGAN aims to improve the accuracy and quality of recommendations provided to users. Moreover, the proposed STRGAN model employs negative sampling techniques to ensure that the generated recommendations align with the real data. To evaluate the effectiveness of STRGAN, extensive experiments were conducted on the real world dataset FilmTrust. The empirical results demonstrate that STRGAN outperforms other GAN-based models in various evaluation metrics, such as precision, recall, normalized discounted cumulative gain (NDCG), and mean reciprocal rank (MRR). STRGAN offers a robust and efficient solution for personalized recommendation tasks. The results of experiment support the efficacy of STRGAN, indicating its potential to significantly improve recommendation accuracy in real-world applications.

Objective: The impact of augmented reality (AR) exercise apps on health-related behaviors and outcomes remains underexplored. Grounded in affordance theory, this study investigates how the utilitarian, hedonic, and social affordances of these apps influence user engagement, ultimately affecting continuance intention and subjective well-being in China. Methods: Data were collected via an online survey of 399 users of AR fitness apps in China, recruited through the professional research platform Credamo.com . All constructs were measured with validated scales adapted to the AR exercise context. The hypothesized relationships were tested using structural equation modeling (SEM). Results: The results indicate that utilitarian (environmental embedding, visibility), hedonic (playfulness), and social (competition) affordances are positively associated with user engagement. This engagement, in turn, positively influences users’ subjective well-being and their intention to continue using AR exercise apps. Conclusion: This study extends affordance theory by identifying the affordances in AR fitness apps and validating user engagement as the critical underlying mechanism. The findings suggest that designers should prioritize these affordances and health policymakers should leverage them to develop effective, AR-driven digital sports initiatives.