Search results

The purpose of this paper is to perform a comparative analysis of the delay associated in running two real-time machine learning-based applications, namely, a video quality assessment (VQA) and a phishing detection application by using the edge, fog and cloud computing paradigms.

The VQA algorithm was developed using Android Studio and run on a mobile phone for the edge paradigm. For the fog paradigm, it was hosted on a Java server and for the cloud paradigm on the IBM and Firebase clouds. The phishing detection algorithm was embedded into a browser extension for the edge paradigm. For the fog paradigm, it was hosted on a Node.js server and for the cloud paradigm on Firebase.

For the VQA algorithm, the edge paradigm had the highest response time while the cloud paradigm had the lowest, as the algorithm was computationally intensive. For the phishing detection algorithm, the edge paradigm had the lowest response time, and the cloud paradigm had the highest, as the algorithm had a low computational complexity. Since the determining factor for the response time was the latency, the edge paradigm provided the smallest delay as all processing were local.

The main limitation of this work is that the experiments were performed on a small scale due to time and budget constraints.

A detailed analysis with real applications has been provided to show how the complexity of an application can determine the best computing paradigm on which it can be deployed.

The use of over-the-top (OTT) platforms grew substantially after the declaration of the COVID-19 pandemic in 2020. With the pandemic receding, there is a concern that users may not continue with their subscriptions. To counter this, OTT service providers must strategize proactively to retain and acquire new users once the pandemic abates. Positing that understanding the consumption values that users ascribe to OTT platform usage can provide useful customer retention insights, the purpose of this paper is to use the theory of consumption value (TCV) to study the values that users derived from their use of OTT following the onset of the pandemic.

The mixed-method approach is used to collect qualitative and quantitative data. Analysis of qualitative responses collected through interviews of 12 current OTT platform users helped identify two categories of OTT platform-specific values: attribute-level and benefit-based. Next, the study examined the association of values thus identified with one another, as well as with continued intentions to use OTT platforms, by analyzing data collected from 371 existing users.

The findings indicated that functional value quality and social value, representing the attribute-level values, were positively associated with two benefit-based values – functional value price and emotional value (EMV). Next, EMV was not only associated with intentions but also partially mediated the association of attribute-level values with intentions. Premium subscription purchased and increased viewing time were confirmed to have moderating effects on the association between attribute-level and benefit-based values.

The study is amongst the foremost research initiatives to examine consumption values derived from OTT platform usage after the onset of the pandemic. Its novelty also comes from its identifying OTT platform-specific consumption values for the first time and adding a new dimension to the TCV by examining the interplay of these values in the OTT platform context.

This study examines the relationship between streamers' product descriptions, customer comments and online sales and focuses on the moderating effect of streamer–viewer relationship strength.

Between June 2021 and April 2022, the structured data of 965 livestreaming and unstructured text data of 42,956,147 characters from two major live-streaming platforms were collected for the study. Text analysis and regression analysis methods were employed for data analysis.

First, the authors' analysis reveals an inverted U-shaped relationship between comment length and product sales. Notably, comment volume and comment emotion positively influence product sales. Furthermore, the semantic richness, emotion and readability of streamers' product descriptions also positively influence product sales. Secondly, the authors find that the strength of streamer–viewer relationship weakens the positive effects of comment volume and comment emotion without moderating the inverted U-shaped effect of comment length. Lastly, the strength of streamer–viewer relationship also diminishes the positive effects of emotion, semantics and readability of streamers' product descriptions on product sales.

This study is the first to concurrently examine the direct and interactive effects of user-generated content (UGC) and marketer-generated content (MGC) on consumer purchase behaviors in livestreaming e-commerce, offering a novel perspective on individual decision-making and cue utilization in the social retail context.

Despite to useful relevance to better understand how group-level identity develops, few studies have explored the identity theory in the esports field and, in particular, considering the impact of a fan's role identity. The current study aims to explore esports fan role-identity vis-à-vis the relationship with the sponsor and the sponsee so as to understand the effects on their behavioral intentions.

Using a sample of 356 esports fans who attended the 2021 FPF eFootball Open Challenge, a Confirmatory Factor Analysis (CFA) analyzed the psychometric properties of the constructs and a subsequent Structural Equation Modeling (SEM) examined the effects of fan identity on two types of behavioral intentions and sponsor–sponsee relationship.

Results indicate that fans who highly identify with esports have the highest attachment to the event and tend toward having a positive word-of-mouth intention. Esports fans who have a higher brand identification reported a positive attitude toward the event's sponsor brand and tend to purchase its products. Moreover, the study findings also provide evidence of the bidirectional interaction between the way in which fans attach with the esports event and its sponsor brand, leading to greater reciprocity in their identity formation.

This study helps to understand how the fan identity process can enhance its fate and develop mutually, building role overlapping identity in the esports sponsor–sponsee relationship. Complementarily, it supports of how the marketeers and managers must analyze the importance of being a fan to the individual in order to understand how its self-identity can shape the future behavior.

In the COVID-19 era, sign language (SL) translation has gained attention in online learning, which evaluates the physical gestures of each student and bridges the communication gap between dysphonia and hearing people. The purpose of this paper is to devote the alignment between SL sequence and nature language sequence with high translation performance.

SL can be characterized as joint/bone location information in two-dimensional space over time, forming skeleton sequences. To encode joint, bone and their motion information, we propose a multistream hierarchy network (MHN) along with a vocab prediction network (VPN) and a joint network (JN) with the recurrent neural network transducer. The JN is used to concatenate the sequences encoded by the MHN and VPN and learn their sequence alignments.

We verify the effectiveness of the proposed approach and provide experimental results on three large-scale datasets, which show that translation accuracy is 94.96, 54.52, and 92.88 per cent, and the inference time is 18 and 1.7 times faster than listen-attend-spell network (LAS) and visual hierarchy to lexical sequence network (H2SNet) , respectively.

In this paper, we propose a novel framework that can fuse multimodal input (i.e. joint, bone and their motion stream) and align input streams with nature language. Moreover, the provided framework is improved by the different properties of MHN, VPN and JN. Experimental results on the three datasets demonstrate that our approaches outperform the state-of-the-art methods in terms of translation accuracy and speed.

This study aims to identify leather type and authenticity through optical coherence tomography.

Optical coherence tomography images taken from genuine and faux leather samples were used to create an image dataset, and automated machine learning algorithms were also used to distinguish leather types.

The optical coherence tomography scan results in a different image based on leather type. This information was used to determine the leather type correctly by optical coherence tomography and automatic machine learning algorithms. Please note that this system also recognized whether the leather was genuine or synthetic. Hence, this demonstrates that optical coherence tomography and automatic machine learning can be used to distinguish leather type and determine whether it is genuine.

For the first time to the best of the authors' knowledge, spectral-domain optical coherence tomography and automated machine learning algorithms were applied to identify leather authenticity in a noncontact and non-invasive manner. Since this model runs online, it can readily be employed in automated quality monitoring systems in the leather industry. With recent technological progress, optical coherence tomography combined with automated machine learning algorithms will be used more frequently in automatic authentication and identification systems.

This study aims to explore the application of emojis to mood descriptions of fiction. The three goals are investigating whether Cho et al.'s model (2023) is a sound conceptual framework for implementing emojis and mood categories in information systems, mapping 30 mood categories to 115 face emojis and exploring and visualizing the relationships between mood categories based on emojis mapping.

An online survey was distributed to a US public university to recruit adult fiction readers. In total, 64 participants completed the survey.

The results show that the participants distinguished between the three families of fiction mood categories. The three families model is a promising option to improve mood descriptions for fiction. Through mapping emojis to 30 mood categories, the authors identified the most popular emojis for each category, analyzed the relationships between mood categories and examined participants' consensus on mapping.

This study focuses on applying emojis to fiction reading. Emojis were mapped to mood categories by fiction readers. Emoji mapping contributes to the understanding of the relationships between mood categories. Emojis, as graphic mood descriptors, have the potential to complement textual descriptors and enrich mood metadata for fiction.

This study aims to empirically test how sponsored video customization (i.e. the degree to which a sponsored video is customized for a sponsoring brand) affects video shares differently depending on influencer characteristics (i.e. mega influencer and expert influencer) and brand characteristics (i.e. brand establishment and product involvement).

This study uses a unique real-world data set that combines coded variables (e.g. customization) and objective video performance (e.g. sharing) of 365 sponsored videos to test the hypotheses. A negative binomial model is used to analyze the data set.

This study finds that the effect of video customization on video shares varies across contexts. Video customization positively affects shares if they are made for well-established brands and high-involvement products but negatively influences shares if they are produced by mega and expert influencers.

This study extends the influencer marketing literature by focusing on a new media modality – sponsored video. Drawing on the multiple inference model and the persuasion knowledge theory, this study teases out different conditions under which video customization is more or less likely to foster audience engagement, which both influencers and brands care about. The chosen research setting may limit the generalizability of the findings of this study.

The findings suggest that mega and expert influencers need to consider if their endorsement would backfire on a highly customized video. Brands that aim to engage customers with highly-customized videos should gauge their decision by taking into consideration their years of establishment and product involvement. For video-sharing platforms, especially those that are planning to expand their businesses to include “matching-making services” for brands and influencers, the findings provide theory-based guidance on optimizing such matches.

This paper fulfills an urgent research need to study how brands and influencers should produce sponsored videos to achieve optimal outcomes.

The principal research objective entails examining the nexus between green supplier selection and green performance while scrutinizing the moderating role of governance mechanisms, specifically process control and outcome control, in shaping this association.

To assess our hypotheses, this study obtained data from Chinese manufacturing sectors and utilized regression analysis on a dataset consisting of 295 samples.

This study enriches the sustainable supply chain management literature by emphasizing the influence of green supplier selection on a firm’s green performance and the moderating effects of outcome and process control, offering practical insights for industry professionals.

This study enriches the sustainable supply chain management literature by emphasizing the influence of supplier selection on a firm’s environmental performance and the moderating effects of outcome and process control, offering practical insights for industry professionals.

The paper aims at assessing the role of the different stages of the employment process in gauging workers' willingness to upskill themselves at the end of a seasonal employment contract by investing in further training.

The paper analyses data from a dedicated survey administered to a sample of seasonal employees. Through a regression analysis it explores the different stages of the employment process (job search, selection on the job activities), making a distinction between monetary and nonmonetary components of the investment in training.

Results show that all stages matter, but they do not have the same importance. Ex-ante motivations and work experience, notably the level of perceived workload and organizational commitment, are the main factors affecting workers' willingness to acquire industry-specific skills through training.

So far, the literature has extensively dealt with the poor levels of training in seasonal employers, but it did not analyse worker’s willingness to invest in training over the different stages of the worker experience. This paper fills this gap by separately testing the relative importance of such stages and identifying the most important phases of the employment process.