Aim/Purpose: This study aims to explore students’ digital entrepreneurial intention (DEI) in Kuwait. Specifically, the aim is twofold: (i) to identify and examine the factors influencing and predicting students’ DEI, and (ii) to validate a model of DEI. Background: The advent of modern digital technologies has provided entrepreneurs with many opportunities to establish and expand their firms through online platforms. Although the existing literature on DEI has explored various factors, certain factors that could be linked to DEI have been neglected, and others have not been given sufficient attention. Nonetheless, there has been little research on students’ DEI, particularly in Kuwait. Methodology: To fulfill the research’s aims, a study was conducted using a quantitative method (a survey of 305 students at a non-profit university in Kuwait). Contribution: This study aimed to fill the research gap on the limited DEI research among Kuwait’s students. Several recommendations were suggested to improve the DEI among students in Kuwait. Findings: The study identified five factors that could influence an individual’s intention to engage in digital entrepreneurship. These factors include self-perceived creativity, social media use, risk-taking and opportunity recognition, digital entrepreneurship knowledge, and entrepreneurial self-perceived confidence. Significant solid correlations were between all five identified factors and DEI. However, only self-perceived creativity and entrepreneurial self-perceived confidence were identified as significant positive predictors of DEI among undergraduates in Kuwait. Nevertheless, the main contributor to this intention was the students’ self-perceived confidence as entrepreneurs. Recommendation for Researchers: Researchers should conduct further longitudinal studies to understand better the dynamic nature of DEI and execution. Future Research: Additional research is required to utilize probability sampling approaches and increase the sample size for more generalizable findings.

Aim/Purpose: The paper discusses various factors contributing to disagreements, such as differing experiences, perspectives, and historical narratives, leading to disagreements within families and societies. It explores how beliefs, values, and biases feed into disagreements, with confirmation bias affecting decision-making and the media. Cultural values also play a role, showcasing conflicts between meritocracy and inclusivity in ethical decision-making. Haidt's Moral Foundations Theory highlights differences in value priorities between Western and Eastern societies. The impact of Western values like rationalism, freedom, and tolerance, under threat from Marxist illiberalism on campuses, is dis-cussed. The text also delves into disinformation, emotions in warfare, and the use of fake information and images for propaganda purposes. The need for diligent reporting to avoid spreading disinformation is emphasized, given its potential to create misconceptions and harm diplomatic relations.

Aim/Purpose: The aim of the present study is to investigate how pre-game superstition and anxiety can drive the consumption and purchase of sports products and objects by triathletes. Methodology: We tested our hypotheses via a cross-sectional study on a sample of N=124 triathletes. Contribution: The originality of our work stands in the provision of empirical evidence on the role of superstition and anxiety in characterized consumer decision-making of triathletes. Theoretically and practically, our results can extend our knowledge of the role of cognitive factors in consumer behaviors among athletes. Findings: The results of the Structural Equation Modelling provided evidence of our hypothesized relationship between pre-game anxiety and superstition, and cognitive biases. Pre-game anxiety increases the level of incidence of specific cognitive biases characterized by intuitive and implicit thinking, while superstition leads to more rational and personal cognitive biases, which affect their purchasing of sports products before games and competitions.

Aim/Purpose: In this paper we propose a framework identifying many of the unintended consequences of information technology and posit that the increased complexity brought about by IT is a proximate cause for these negative effects. Background: Builds upon the three-world model that has been evolving within the informing science transdiscipline. Methodology: We separate complexity into three categories: experienced complexity, intrinsic complexity, and extrinsic complexity. With the complexity cycle in mind, we consider how increasing complexity of all three forms can lead to unintended consequences at the individual, task and system levels. Examples of these consequences are discussed at the individual level (e.g., deskilling, barriers to advancement), the task level (e.g., perpetuation of past practices), as well as broader consequences that may result from the need to function in an environment that is more extrinsically complex (e.g., erosion of predictable causality, shortened time horizons, inequality, tribalism). We conclude by reflecting on the implications of attempting to manage or limit increases of complexity. Contribution: Shows how many unintended consequences of IT could be attributed to growing complexity. Findings: We find that these three forms of complexity feed into one another resulting in a positive feedback loop that we term the Complexity Cycle. As examples, we analyze ChatGPT, blockchain and quantum computing, through the lens of the complexity cycle, speculating how experienced complexity can lead to greater intrinsic complexity in task performance through the incorporation of IT which, in turn, increases the extrinsic complexity of the economic/technological environment. Recommendations for Practitioners: Consider treating increasing task complexity as an externality that should be considered as new systems are developed and deployed. Recommendation for Researchers: Provides opportunities for empirical investigation of the proposed model. Impact on Society: Systemic risks of complexity are proposed along with some proposals regarding how they might be addressed. Future Research: Empirical investigation of the proposed model and the degree to which cognitive changes created by the proposed complexity cycle are necessarily problematic.

Aim/Purpose: This review paper aims to unveil some underlying machine-learning classification algorithms used for political election predictions and how stack ensembles have been explored. Additionally, it examines the types of datasets available to researchers and presents the results they have achieved. Background: Predicting the outcomes of presidential elections has always been a significant aspect of political systems in numerous countries. Analysts and researchers examining political elections rely on existing datasets from various sources, including tweets, Facebook posts, and so forth to forecast future elections. However, these data sources often struggle to establish a direct correlation between voters and their voting patterns, primarily due to the manual nature of the voting process. Numerous factors influence election outcomes, including ethnicity, voter incentives, and campaign messages. The voting patterns of successors in regions of countries remain uncertain, and the reasons behind such patterns remain ambiguous. Methodology: The study examined a collection of articles obtained from Google Scholar, through search, focusing on the use of ensemble classifiers and machine learning classifiers and their application in predicting political elections through machine learning algorithms. Some specific keywords for the search include “ensemble classifier,” “political election prediction,” and “machine learning”, “stack ensemble”. Contribution: The study provides a broad and deep review of political election predictions through the use of machine learning algorithms and summarizes the major source of the dataset in the said analysis. Findings: Single classifiers have featured greatly in political election predictions, though ensemble classifiers have been used and have proven potent use in the said field is rather low. Recommendation for Researchers: The efficacy of stack classification algorithms can play a significant role in machine learning classification when modelled tactfully and is efficient in handling labelled datasets. however, runtime becomes a hindrance when the dataset grows larger with the increased number of base classifiers forming the stack. Future Research: There is the need to ensure a more comprehensive analysis, alternative data sources rather than depending largely on tweets, and explore ensemble machine learning classifiers in predicting political elections. Also, ensemble classification algorithms have indeed demonstrated superior performance when carefully chosen and combined.

Aim/Purpose: The truism is that leadership fosters or restricts innovation behaviours in organisations, but the extent to which it does depends on the leadership style in practice. This study focuses on one of the contemporary leadership styles, knowledge-oriented leadership [KOL], which has received scant attention in research. In doing so, the contextual factors of psychological safety [PS] and employee voice [EV] were applied to determine how KOL influences are channeled to innovation at the individual level. Methodology: Data were collected from 347 academic staff in public universities in Southern Nigeria and subjected to a partial least square [PLS] analytical procedure for data treatment and hypotheses testing using the SmartPLS 3 software for variance-based structural equation modelling. Contribution: The study formed an integrated research framework that links knowledge-oriented leadership and innovation by accounting for the contextual mechanisms of psychological safety and employee voice. Findings: The PLS results demonstrated that the knowledge-oriented leadership and innovation relationship was positive and significant, and this relationship was partially mediated by two variables, namely, PS and EV. Furthermore, the two mediating variables channeled KOL’s influence on innovation in a sequence. Recommendation for Researchers: Organisations need to consider the practical application of KOL to improve innovation outcomes considerably. By this, leadership training programs should include modules, courses, or topics on KOL to engender the formation of requisite managerial skills. More so, they should consider the criterion of demonstrable KOL abilities for leadership selection and recruitment. As a personal development initiative, managers can attend leadership development programmes as well as obtain certification in knowledge management to improve their KOL abilities. This initiative should be encouraged and supported by organisations. In all, the human resource management framework should be responsive to the dynamics of the knowledge economy regarding leadership. Given that PS and EV function as mediators, organisations should actively cultivate an environment enabling interpersonal risky behaviours founded on trust, respect, and cooperation and encourage/support employees who demonstrate such behaviour accordingly. In this line, they should create and sustain a supportive environment that positively reinforces voice decisions and behaviours. Future Research: The study only determined the links between KOL, PS, EV, and innovation in public universities in Southern Nigeria. Other studies may examine the linkages in other knowledge-intensive organisations as well as expand the geographic scope to make for better generality of findings. Future studies should look at other underlying mechanisms that can affect the KOL-innovation relationship, such as psychological capital, work engagement, work commitment, etc. The role of moderators can be identified and introduced to this integrative framework to demonstrate the conditions affecting the linkages.

Table of Contents for Volume 27 of Informing Science: The International Journal of an Emerging Transdiscipline, 2024

Labor pain and its relieving methods is one of the anxieties of mothers having a great impact on the quality of care during delivery as well as the patients' satisfaction. The propensity of using non-medicinal pain relief methods is increasing. The present study aimed to compare the effect of Acupressure at two GB-21 and SP06 points on the severity of labor pain. In this quasi-experimental single blind study started on December 2010 and ended on June 2011 in which 150 primiparous women were divided into three groups of Acupressure at GB-21 point, Acupressure at SP-6 point and control group. The intervention was carried out for 20 min at 3-4 and 20 min at 7-8 cm dilatation of Cervix. The pain severity was measured by Visual Analog Scale before and immediately, 30 and 60 min after the intervention. Then, the data were statistically analyzed. No significant difference was found among the 3 groups regarding the pain severity before the intervention. However, the pain severity it was reduced at 3-4 and 7-8 cm dilatation immediately, 30 and 60 min after the intervention in the two intervention groups compared to the control group (p<0.001). Nonetheless, no statistically significant difference was observed between the two intervention groups (p = 0.93). The results of the study showed that application of Acupressure at two GB-21 and SP-6 points was effective in the reduction of the severity of labor pain. Therefore, further studies are recommended to be performed on the application of Acupressure together with non-medicinal methods.

Friday, September 27 | 9:30 a.m. (PT) | Online Event Details and Recoding Here  Join us for Race, Rights, and Innovation: Cultivating Equity in the Digital World, a thought-provoking event exploring the intersection of race, technology, and legal frameworks. We’ll delve into the historical treatment of minority creators in copyright ...

The post 2024 Fall Symposium — Race, Rights, and Innovation: Cultivating Equity in the Digital World appeared first on Berkeley Technology Law Journal.

This article is part of the 2024 BCLT-BTLJ-CMTL Symposium.  Angela R. Riley When my tribe, the Citizen Potawatomi Nation of Oklahoma (CPN), established an Eagle Aviary to protect and care for injured eagles that could no longer survive in the wild, it did so with a few goals in mind. ...

The post Tribal Self-Determination and the Protection of Cultural Property appeared first on Berkeley Technology Law Journal.

This article is part of the 2024 BCLT-BTLJ-CMTL Symposium.  Leo Yu The national security concerns surrounding TikTok appear straightforward: it is China. To many policymakers and scholars, the mere connection to China warrants severe measures, including either divestment to an American firm or a complete shutdown. What renders China’s involvement ...

The post TikTok and the Control over the Means of Production in the Fourth Industrial Revolution appeared first on Berkeley Technology Law Journal.

By Barbara Rasin, J.D. Candidate, 2027 In Anderson v. TikTok, decided in in late summer 2024, the Third Circuit Court of Appeals held that TikTok’s “For You Page” algorithm was sufficiently creative to bar its protection under §230 of the Communications Decency Act (CDA). This is a significant step towards ...

The post Expanding TikTok’s Liability for the “For You Page” appeared first on Berkeley Technology Law Journal.

[Meg O’Neill] 00:08 Hello and welcome to the Berkeley Technology Law Journal podcast. My name is Meg O’Neill and I am one of the editors of the podcast. Today we are excited to share with you a conversation between Berkeley Law LLM student Franco Dellafiori, and Professor Bertrall Ross. Professor ...

The post Berkeley Technology Law Journal Podcast: Will ChatGPT Tell Me How to Vote? Democracy & AI with Professor Bertrall Ross appeared first on Berkeley Technology Law Journal.

This paper proposes an efficient solution for personalised web directories recommendation using fast FCM+DQN. At first, web directory usage file obtained from given dataset is fed into the accretion matrix computation module, where visitor chain matrix, visitor chain binary matrix, directory chain matrix and directory chain binary matrix are formulated. In this, directory grouping is accomplished based on fast FCM and matching among query and group is conducted based on Kumar Hassebrook and Kulczynski similarity. The user preferred directory is restored at this stage and at last, personalised web directories are recommended to the visitors by means of DQN. The proposed approach has received superior results with respect to maximum accuracy of 0.910, minimum mean squared error (MSE) of 0.0206 and root mean squared error (RMSE) of 0.144. Although the system offered magnificent outcomes, it failed to order web directories in the form of highly, medium and low interested directories.

Mental illnesses are common among college students as well as their non-student peers, and the number and severity of these problems are increasing. It can be difficult to identify people suffering from mental illness and get the help they need early. So in this paper, the SqueezeR_MobileNet method is proposed. It performs feature fusion and early mental health prediction. Initially, outliers in the input data are detected and removed. After that, using missing data imputation and Z-score normalisation the pre-processing phase is executed. Next to this, for feature fusion, a combination of the Soergel metric and deep Kronecker network (DKN) is used. By utilising bootstrapping data augmentation is performed. Finally, early mental health prediction is done using SqueezeR_MobileNet, which is the incorporation of residual SqueezeNet and MobileNet. The devised approach has reached the highest specificity of 0.937, accuracy of 0.911 and sensitivity of 0.907.

This paper presents a novel deep learning technique called quantum dilated convolutional neural network-DenseNet (Q-DenseNet) for prediction of heart disease in spark framework. At first, the input data taken from the database is allowed for data partitioning using fast fuzzy C-means clustering (FFCM). The partitioned data is fed into spark framework, where pre-processed by missing data imputation and quantile normalisation. The pre-processed data is further allowed for selection of suitable features. Then, the selected features from the slave nodes are merged and fed into master node. The Q-DenseNet is used in master node for the prediction of heart disease. The performance improvement of the designed Q-DenseNet model is validated by comparing with traditional prediction models. Here, the Q-DenseNet method achieved superior performance with maximum of 92.65% specificity, 91.74% sensitivity, and 90.15% accuracy.

An optimisation-based decision-making support is proposed in this study in the form of fuzzy-probabilistic programming, which can be used to solve integrated order allocation, production planning, and inventory management problems in fuzzy and probabilistic uncertain environments. The problem was modelled in an uncertain mathematical optimisation model with two objectives: maximising the expectation of production volume and minimising the expectation of total operational cost subject to demand and other constraints. The model belongs to fuzzy-probabilistic bi-objective integer linear programming, and the generalised reduced gradient method combined with the branch-and-bound algorithm was utilised to solve the derived model. Numerical simulations were performed to illustrate how the optimal decision was formulated. The results showed that the proposed decision-making support was successful in providing the optimal decision with the maximum expectation of the production volume and minimum expectation of the total operational cost. Therefore, the approach can be implemented by decision-makers in manufacturing companies.

COVID-19 created a surge in e-commerce usage, leading to fierce channel competition between the manufacturer's online sales and the offline retailer. Hence, the imperative need for effective and innovative optimisation strategies to mitigate channel competition. Manufacturer-coupons are widely practiced in market, yet research on the importance they play in coordinating channel competition to achieve optimisation in channel distributions is scarce. This research addresses this gap by examining the effectiveness of manufacturer-coupons on the coordination of the manufacturer's online sales and offline retailer's sales. The findings indicate that issuing a manufacturer-coupon to the customers who buy from the offline retailer reduces the competition in the different channel distributions, but cost sharing of the retailer coupon is a better strategy. We thus examine if profit sharing is an effective strategy to facilitate the use of manufacturer-coupon in the market. After comparing different scenarios, we conclude that advanced profit-sharing can be effective in making manufacturer-coupon prevalent in the market and thus alleviate channel competition effectively.

Data envelopment analysis (DEA) is widely used in various fields and for various models. Inverted data envelopment analysis (inverted DEA) is an extended model of DEA. Regression analysis (RA) is a statistical process for estimating the relationships among variables based on the model of averaged image. There are no essential relations among DEA and RA and inverted DEA. We creatively combine DEA, RA and inverted DEA to propose a new model: DEA-RA-Inverted DEA model. The model realises the efficiency estimation and evaluation through a discussion of the residual variables and the residual ratio coefficients. In addition, we will demonstrate the effectiveness of the model by applying it to efficiency estimation and evaluation of 16 Chinese logistics enterprises.

This study addresses a critical chain project scheduling (CCPS) problem regarding the feeding buffer. The main contribution of this study lies in determining the critical chain when the feeding buffer is considered along with the project buffer, a less addressed issue in the critical chain literature. Using a mixed-integer nonlinear programming (MINLP) model, the critical chain of a project with no break-down and no overflow is found. Moreover, the impact of the feeding buffer on the criticality of activities is discussed. The problem is solved using the Lingo software package for validation in small-sized instances. Since the CCPS is known as an NP-hard problem, a genetic algorithm (GA) is also designed to solve large-scale instances. The algorithm's performance is confirmed using various project scheduling library test problems. Sensitivity analysis is implemented based on some crucial parameters, and the critical chain is analysed after conducting several experiments. It is shown how considering the feeding buffer makes different critical chains and how shortlisting activities and resources are optimally managed.

This paper studies a dual-channel supply chain consisting of one risk-averse manufacturer and one risk-averse retailer with stochastic demand. Herein, the manufacturer provides value-added services to enhance channel demand. First, the optimal pricing and service decisions of the channel members are investigated under different settings, i.e., the cooperative game, Bertrand game, and manufacturer Stackelberg (MS) game models. Second, the effects of channel members' risk aversion on optimal channel prices and expected utilities are analysed under the assumption that the manufacturer service is a decision variable and an exogenous variable, respectively. Third, sensitivity analysis and numerical simulation are performed to verify our propositions consistently and seek more managerial implications. The findings suggest that the manufacturer's value-added services in their direct channel will improve the direct price while decreasing the retail price. Consumers' channel loyalty degree has a great influence on the optimal price decisions and the performance of the channel members. The direct price increases while the retail price decreases in the manufacturer's value-added services. The retailer's risk aversion has a greater influence on price decisions than that of the manufacturer.

Lihan TONG,Weijia LI,Qingxia YANG,Liyuan CHEN,Peng CHEN, Vol.E107-D, No.11, pp.1472-1475
We present Ksformer, utilizing Multi-scale Key-select Routing Attention (MKRA) for intelligent selection of key areas through multi-channel, multi-scale windows with a top-k operator, and Lightweight Frequency Processing Module (LFPM) to enhance high-frequency features, outperforming other dehazing methods in tests.
Publication Date: 2024/11/01

Shuoyan LIU,Chao LI,Yuxin LIU,Yanqiu WANG, Vol.E107-D, No.11, pp.1468-1471
Escalators are an indispensable facility in public places. While they can provide convenience to people, abnormal accidents can lead to serious consequences. Yolo is a function that detects human behavior in real time. However, the model exhibits low accuracy and a high miss rate for small targets. To this end, this paper proposes the Small Target High Performance YOLO (SH-YOLO) model to detect abnormal behavior in escalators. The SH-YOLO model first enhances the backbone network through attention mechanisms. Subsequently, a small target detection layer is incorporated in order to enhance detection of key points for small objects. Finally, the conv and the SPPF are replaced with a Region Dynamic Perception Depth Separable Conv (DR-DP-Conv) and Atrous Spatial Pyramid Pooling (ASPP), respectively. The experimental results demonstrate that the proposed model is capable of accurately and robustly detecting anomalies in the real-world escalator scene.
Publication Date: 2024/11/01

Takahito YOSHIDA,Takaharu YAGUCHI,Takashi MATSUBARA, Vol.E107-D, No.11, pp.1458-1462
Accurately simulating physical systems is essential in various fields. In recent years, deep learning has been used to automatically build models of such systems by learning from data. One such method is the neural ordinary differential equation (neural ODE), which treats the output of a neural network as the time derivative of the system states. However, while this and related methods have shown promise, their training strategies still require further development. Inspired by error analysis techniques in numerical analysis while replacing numerical errors with modeling errors, we propose the error-analytic strategy to address this issue. Therefore, our strategy can capture long-term errors and thus improve the accuracy of long-term predictions.
Publication Date: 2024/11/01

Nat PAVASANT,Takashi MORITA,Masayuki NUMAO,Ken-ichi FUKUI, Vol.E107-D, No.11, pp.1453-1457
We proposed a procedure to pre-process data used in a vector autoregressive (VAR) modeling of a temporal point process by using kernel density estimation. Vector autoregressive modeling of point-process data, for example, is being used for causality inference. The VAR model discretizes the timeline into small windows, and creates a time series by the presence of events in each window, and then models the presence of an event at the next time step by its history. The problem is that to get a longer history with high temporal resolution required a large number of windows, and thus, model parameters. We proposed the local density estimation procedure, which, instead of using the binary presence as the input to the model, performed kernel density estimation of the event history, and discretized the estimation to be used as the input. This allowed us to reduce the number of model parameters, especially in sparse data. Our experiment on a sparse Poisson process showed that this procedure vastly increases model prediction performance.
Publication Date: 2024/11/01

Keitaro NAKASAI,Shin KOMEDA,Masateru TSUNODA,Masayuki KASHIMA, Vol.E107-D, No.11, pp.1444-1448
To automatically measure the mental workload of developers, existing studies have used biometric measures such as brain waves and the heart rate. However, developers are often required to equip certain devices when measuring them, and can therefore be physically burdened. In this study, we evaluated the feasibility of non-contact biometric measures based on the nasal skin temperature (NST). In the experiment, the proposed biometric measures were more accurate than non-biometric measures.
Publication Date: 2024/11/01

Qinghua WU,Weitong LI, Vol.E107-D, No.11, pp.1422-1432
Multi-focus image fusion involves combining partially focused images of the same scene to create an all-in-focus image. Aiming at the problems of existing multi-focus image fusion algorithms that the benchmark image is difficult to obtain and the convolutional neural network focuses too much on the local region, a fusion algorithm that combines local and global feature encoding is proposed. Initially, we devise two self-supervised image reconstruction tasks and train an encoder-decoder network through multi-task learning. Subsequently, within the encoder, we merge the dense connection module with the PS-ViT module, enabling the network to utilize local and global information during feature extraction. Finally, to enhance the overall efficiency of the model, distinct loss functions are applied to each task. To preserve the more robust features from the original images, spatial frequency is employed during the fusion stage to obtain the feature map of the fused image. Experimental results demonstrate that, in comparison to twelve other prominent algorithms, our method exhibits good fusion performance in objective evaluation. Ten of the selected twelve evaluation metrics show an improvement of more than 0.28%. Additionally, it presents superior visual effects subjectively.
Publication Date: 2024/11/01

Naoya NEZU,Hiroshi YAMADA, Vol.E107-D, No.11, pp.1408-1421
Modern memory devices such as DRAM are prone to errors that occur because of unintended bit flips during their operation. Since memory errors severely impact in-memory key-value stores (KVSes), software mechanisms for hardening them against memory errors are being explored. However, it is hard to efficiently test the memory error handling code due to its characteristics: the code is event-driven, the handlers depend on the memory object, and in-memory KVSes manage various objects in huge memory space. This paper presents MemFI that supports runtime tests for the memory error handlers of in-memory KVSes. Our approach performs the software fault injection of memory errors at the memory object level to trigger the target handler while smoothly carrying out tests on the same running state. To show the effectiveness of MemFI, we integrate error handling mechanisms into a real-world in-memory KVS, memcached 1.6.9 and Redis 6.2.7, and check their behavior using the MemFI prototypes. The results show that the MemFI-based runtime test allows us to check the behavior of the error handling mechanisms. We also show its efficiency by comparing it to other fault injection approaches based on a trial model.
Publication Date: 2024/11/01

Yuan LI,Tingting HU,Ryuji FUCHIKAMI,Takeshi IKENAGA, Vol.E107-D, No.11, pp.1396-1407
1 millisecond (1-ms) vision systems are gaining increasing attention in diverse fields like factory automation and robotics, as the ultra-low delay ensures seamless and timely responses. Superpixel segmentation is a pivotal preprocessing to reduce the number of image primitives for subsequent processing. Recently, there has been a growing emphasis on leveraging deep network-based algorithms to pursue superior performance and better integration into other deep network tasks. Superpixel Sampling Network (SSN) employs a deep network for feature generation and employs differentiable SLIC for superpixel generation. SSN achieves high performance with a small number of parameters. However, implementing SSN on FPGAs for ultra-low delay faces challenges due to the final layer’s aggregation of intermediate results. To address this limitation, this paper proposes an aggregated to pipelined structure for FPGA implementation. The final layer is decomposed into individual final layers for each intermediate result. This architectural adjustment eliminates the need for memory to store intermediate results. Concurrently, the proposed structure leverages decomposed layers to facilitate a pipelined structure with pixel streaming input to achieve ultra-low latency. To cooperate with the pipelined structure, layer-partitioned memory architecture is proposed. Each final layer has dedicated memory for storing superpixel center information, allowing values to be read and calculated from memory without conflicts. Calculation results of each final layer are accumulated, and the result of each pixel is obtained as the stream reaches the last layer. Evaluation results demonstrate that boundary recall and under-segmentation error remain comparable to SSN, with an average label consistency improvement of 0.035 over SSN. From a hardware performance perspective, the proposed system processes 1000 FPS images with a delay of 0.947 ms/frame.
Publication Date: 2024/11/01

Zhigang WU,Yaohui ZHU, Vol.E107-D, No.11, pp.1385-1395
This article focuses on improving the BiSeNet v2 bilateral branch image segmentation network structure, enhancing its learning ability for spatial details and overall image segmentation accuracy. A modified network called “BiconvNet” is proposed. Firstly, to extract shallow spatial details more effectively, a parallel concatenated strip and dilated (PCSD) convolution module is proposed and used to extract local features and surrounding contextual features in the detail branch. Continuing on, the semantic branch is reconstructed using the lightweight capability of depth separable convolution and high performance of ConvNet, in order to enable more efficient learning of deep advanced semantic features. Finally, fine-tuning is performed on the bilateral guidance aggregation layer of BiSeNet v2, enabling better fusion of the feature maps output by the detail branch and semantic branch. The experimental part discusses the contribution of stripe convolution and different sizes of empty convolution to image segmentation accuracy, and compares them with common convolutions such as Conv2d convolution, CG convolution and CCA convolution. The experiment proves that the PCSD convolution module proposed in this paper has the highest segmentation accuracy in all categories of the Cityscapes dataset compared with common convolutions. BiConvNet achieved a 9.39% accuracy improvement over the BiSeNet v2 network, with only a slight increase of 1.18M in model parameters. A mIoU accuracy of 68.75% was achieved on the validation set. Furthermore, through comparative experiments with commonly used autonomous driving image segmentation algorithms in recent years, BiConvNet demonstrates strong competitive advantages in segmentation accuracy on the Cityscapes and BDD100K datasets.
Publication Date: 2024/11/01

One of the main tasks in tax administration is debt management. The main goal of this function is tax due collection. Statements are processed in order to select strategies to use in the debt management process to optimise the debt collection process. This work proposes to carry out a data mining process to predict debts of taxpayers with high probability of non-payment. The data mining process identifies high-risk debts using a survival analysis on a dataset from a tax administration. Three groups of tax debtors with similar payment behaviour were identified and a success rate of up to 90% was reached in estimating the payment time of taxpayers. The concordance index (C-index) was used to determine the performance of the constructed model. The highest prediction rate reached was 90.37% corresponding to the third group.

Product innovation has become a crucial part of the sustainability of the coconut agro-industry in Indonesia, covering upstream and downstream sides. To overcome this challenge, it is necessary to create several model stages using a hybrid method that combines machine learning based on multiple criteria decision making and mass balance analysis. The research case study was conducted in Tembilahan district, Riau province, Indonesia, one of the primary coconut producers in Indonesia. The analysis results showed that potential products for domestic customers included coconut milk, coconut cooking oil, coconut chips, coconut jelly, coconut sugar, and virgin coconut oil. Furthermore, considering the experts, the most potential product to be developed was coconut sugar with a weight of 0.26. Prediction of coconut sugar demand reached 13,996,607 tons/year, requiring coconut sap as a raw material up to 97,976,249.

The Republic of India has witnessed an enormous leap in financial transactions after a sudden demonetisation in 2016. The study represents an in-depth analysis of the factors influencing e-wallets usage post-COVID situation covering the National Capital Region. The scientifically collected data were subjected to Pearson's correlation to recognise the correlation amongst the selected e-wallets. The usage of e-wallets is observed mainly during recharge, UPI payments, and utility payments. Through psychometric response and interaction analysis, six factors were selected and examined for data distribution and stable observation using standard deviation and variance coefficient. The coefficient of variance for six factors was observed ≤ 1. The weight of the factors noted to be secured way (0.184), to take advantage of cashback (0.182), low risk of theft (0.169), fast service (0.1689), ease to use (0.156), and saves time (0.139) using principal component eigenvectors analysis. Freecharge and Tez wallets reveal a maximum 99.2% correlation.

Research growth in organisational behaviour research, has increased the importance of paying attention to anti-citizenship behaviours. The current research with the aim of quantitative combination, has examined the results of research in effect of underlying factors of organisational anti-citizenship behaviours using meta-analysis method and CMA2 software and 55 articles during the time period of 2000-2020. The results showed a positive significant link between underlying factors of organisational anti-citizenship behaviours and occurrence of these behaviours and this influence was 0.389, 0.338, 0514 and 0.498 (structural, organisational, managerial, employment and professional and socio-economic and cultural factors). The level of connection found relating to each four occurrences is '68 links, 49 links, 93 links and 71 links'. Findings indicate that minute attention has been paid to organisational anti-citizenship behaviours, especially to job and professional factors in research works. Research should be conducted to control and manage these behaviours more purposefully in organisations.

In Italy, the reimbursement provided to healthcare organisations for medical and surgical procedures is based on the diagnosis related group weight (DRGW), which is an increasing function of the complexity of the procedures. This makes the reimbursement an upper unlimited function. This model does not include the relation of the volume with the complexity. The paper proposes a mathematical model for the optimisation of the reimbursement by determining the optimal mix of volume/procedure, considering the relation volume/complexity and DRGW/complexity. The decreasing, linear, and increasing returns to scale have been defined, and the optimal solution found. The comparison of the model with the traditional approach shows that the proposed model helps the healthcare system to discern the quantity of the reimbursement to provide to health organisations, while the traditional approach, neglecting the relation between the volume and the complexity, can result in an overestimation of the reimbursement.

Underpinned by stakeholder and agency theories, this study adopts a qualitative multiple-case study approach to explore and analyse various stakeholder interests and how they affect digitalisation in the health sector of a developing country (DC). The study's findings revealed that four key stakeholder interests - political, regulatory, leadership, and operational - affect digitalisation in the health sector of DCs. Further, the study found that operational and leadership interests were emergent and were triggered by some digitalisation initiatives, which included, inter alia, the use of new eHealth software and the COVID-19 vaccination exercise, which established new structures and worked better through digitalisation. Conversely, political and regulatory interests were found to be relatively enduring since they existed throughout the pre- and post-digitalisation eras. The study also unearthed principal-agent conflicts arising from technological, organisational and regulatory factors that contribute to the paradoxical outcomes of digitalisation in the health sector.

Recent innovations in additive manufacturing (AM) have proven its efficacy for not only the manufacturing industry but also the healthcare industry. Researchers from Cal Poly, San Luis Obispo, and California State University Long Beach are developing a model that will determine the optimal locations for additive manufacturing hubs that can effectively serve both the manufacturing and healthcare industries. This paper will focus on providing an overview of the healthcare industry's unique needs for an AM hub and summarise the specific inputs for the model. The methods used to gather information include extensive literature research on current practices of AM models in healthcare and an inclusive survey of healthcare practitioners. This includes findings on AM's use for surgical planning and training models, the workflow to generate them, sourcing methods, and the AM techniques and materials used. This paper seeks to utilise the information gathered through literature research and surveys to provide guidance for the initial development of an AM hub location model that locates optimal service locations.

Collaboration between various stakeholders is crucial for healthcare digitalisation and eHealth utilisation. Given that valuable outcomes can emerge from collaborative interactions among multiple stakeholders, exploring a quadruple helix (QH) approach to collaboration may be fruitful in involving the public sector, business, citizens, and academia. Therefore, this study aimed to explore stakeholder views on eHealth collaboration from a QH perspective using the grounded theory methodology. First, an inductive qualitative study involving all stakeholders in the QH was conducted. Subsequently, the findings were related to the actor-resource-activity (ARA) model of business relationships. The results emphasise the role of considering diverse perspectives on collaboration because digitalisation and eHealth require teamwork to benefit the end users within various settings. A model depicting the various aspects of the ARA model related to digitalisation in a healthcare QH setting is presented.

Background and Objective: In response to the challenges of poor mapping outcomes and susceptibility to obstacles encountered by indoor mobile vehicles relying solely on pure cameras or pure LiDAR during their movements, this paper proposes an obstacle avoidance method for indoor mobile vehicles that integrates image and LiDAR data, thus achieving obstacle avoidance for mobile vehicles. Materials and Methods: This method combines data from a depth camera and LiDAR, employing the Gmapping SLAM algorithm for environmental mapping, along with the A* algorithm and TEB algorithm for local path planning. In addition, this approach incorporates gesture functionality, which can be used to control the vehicle in certain special scenarios where “pseudo-obstacles” exist. The method utilizes the YOLO V3 algorithm for gesture recognition. Results: This paper merges the maps generated by the depth camera and LiDAR, resulting in a three-dimensional map that is more enriched and better aligned with real-world conditions. Combined with the A* algorithm and TEB algorithm, an optimal route is planned, enabling the mobile vehicles to effectively obtain obstacle information and thus achieve obstacle avoidance. Additionally, the introduced gesture recognition feature, which has been validated, also effectively controls the forward and backward movements of the mobile vehicles, facilitating obstacle avoidance. Conclusion: The experimental platform for the mobile vehicles, which integrates depth camera and LiDAR, built in this study has been validated for real-time obstacle avoidance through path planning in indoor environments. The introduced gesture recognition also effectively enables obstacle avoidance for the mobile vehicles.

A semantically rich framework to enable real-time knowledge extraction from short length semi-structured documents Lavana Elluri 10:30-11:30 Monday, 4 November 2019, ITE346 Knowledge is currently maintained as a large volume of unstructured text data in books, laws, regulations and policies, news and social media, academic and scientific reports, conversation and correspondence, etc. Most of these […]

The post TALK: Real-time knowledge extraction from short semi-structured documents appeared first on UMBC ebiquity.

Ph.D. Dissertation Defense Modeling and Extracting Information about Cybersecurity Events from Text Taneeya Satyapanich 9:30-11:30 Monday, 18 November, 2019, ITE346? People now rely on the Internet to carry out much of their daily activities such as banking, ordering food, and socializing with their family and friends. The technology facilitates our lives, but also comes with […]

The post Defense: Taneeya Satyapanich, Modeling and Extracting Information about Cybersecurity Events from Text appeared first on UMBC ebiquity.

Reinforcement Quantum Annealing: A Quantum-Assisted Learning Automata Approach We introduce the reinforcement quantum annealing (RQA) scheme in which an intelligent agent interacts with a quantum annealer that plays the stochastic environment role of learning automata and tries to iteratively find better Ising Hamiltonians for the given problem of interest. As a proof-of-concept, we propose a […]

The post Reinforcement Quantum Annealing: A Quantum-Assisted Learning Automata Approach appeared first on UMBC ebiquity.

Results using the reinforcement learning technique on two SAT benchmarks using a D-Wave 2000Q quantum processor showed significantly better solutions with fewer samples compared to the best-known quantum annealing techniques.

The post Paper: Reinforcement Quantum Annealing: A Hybrid Quantum Learning Automata appeared first on UMBC ebiquity.

This paper how to apply dynamic topic models to a set of cybersecurity documents to understand how the concepts found in them are changing over time.

The post paper: Temporal Understanding of Cybersecurity Threats appeared first on UMBC ebiquity.

Basics of the Adtech Ecosystem This guide delves into the intricacies of the adtech ecosystem, an elaborate mesh of platforms and technologies designed to facilitate and enhance the purchase and sale of digital advertising. Within this system, crucial elements such as ad servers, DSPs (Demand Side Platforms), SSPs (Supply Side Platforms), and ad [...]