Processing Time Research Articles

MCI Net: Mamba- Convolutional lightweight self-attention medical image segmentation network.

With the development of deep learning in the field of medical image segmentation, various network segmentation models have been developed. Currently, the most common network models in medical image segmentation can be roughly categorized into pure convolutional networks, Transformer-based networks, and networks combining convolution and Transformer architectures. However, when dealing with complex variations and irregular shapes in medical images, existing networks face issues such as incomplete information extraction, large model parameter sizes, high computational complexity, and long processing times. In contrast, models with lower parameter counts and complexity can efficiently, quickly, and accurately identify lesion areas, significantly reducing diagnosis time and providing valuable time for subsequent treatments. Therefore, this paper proposes a lightweight network named MCI-Net, with only 5.48 M parameters, a computational complexity of 4.41, and a time complexity of just 0.263. By performing linear modeling on sequences, MCI-Net permanently marks effective features and filters out irrelevant information. It efficiently captures local-global information with a small number of channels, reduces the number of parameters, and utilizes attention calculations with exchange value mapping. This achieves model lightweighting and enables thorough interaction of local-global information within the computation, establishing an overall semantic relationship of local-global information. To verify the effectiveness of the MCI-Net network, we conducted comparative experiments with other advanced representative networks on five public datasets: X-ray, Lung, ISIC-2016, ISIC-2018, and capsule endoscopy and gastrointestinal segmentation. We also performed ablation experiments on the first four datasets. The experimental results outperformed the other compared networks, confirming the effectiveness of MCI-Net. This research provides a valuable reference for achieving lightweight, accurate, and high-performance medical image segmentation network models.

Unbiased and reproducible liver MRI-PDFF estimation using a scan protocol-informed deep learning method.

To estimate proton density fat fraction (PDFF) from chemical shift encoded (CSE) MR images using a deep learning (DL)-based method that is precise and robust to different MR scanners and acquisition echo times (TEs). Variable echo times neural network (VET-Net) is a two-stage framework that first estimates nonlinear variables of the CSE-MR signal model, to posteriorly estimate water/fat signal components using the least-squares method. VET-Net incorporates a vector with TEs as an auxiliary input, therefore enabling PDFF calculation with any TE setting. A single-site liver CSE-MRI dataset (188 subjects, 4146 axial slices) was considered, which was split into training (150 subjects), validation (18), and testing (20) subsets. Testing subjects were scanned using several protocols with different TEs, which we then used to measure the PDFF reproducibility coefficient (RDC) at two regions of interest (ROIs): the right posterior and left hepatic lobes. An open-source multi-site and multi-vendor fat-water phantom dataset was also used for PDFF bias assessment. VET-Net showed RDCs of 1.71% and 1.04% on the right posterior and left hepatic lobes, respectively, across different TEs, which was comparable to a reference graph cuts-based method (RDCs = 1.71% and 0.86%). VET-Net also showed a smaller PDFF bias (-0.55%) than graph cuts (0.93%) when tested on a multi-site phantom dataset. Reproducibility (1.94% and 1.59%) and bias (-2.04%) were negatively affected when the auxiliary TE input was not considered. VET-Net provided unbiased and precise PDFF estimations using CSE-MR images from different hardware vendors and different TEs, outperforming conventional DL approaches. Question Reproducibility of liver PDFF DL-based approaches on different scan protocols or manufacturers is not validated. Findings VET-Net showed a PDFF bias of -0.55% on a multi-site phantom dataset, and RDCs of 1.71% and 1.04% at two liver ROIs. Clinical relevance VET-Net provides efficient, in terms of scan and processing times, and unbiased PDFF estimations across different MR scanners and scan protocols, and therefore it can be leveraged to expand the use of MRI-based liver fat quantification to assess hepatic steatosis.

Filtering Useful App Reviews Using Naïve Bayes—Which Naïve Bayes?

App reviews provide crucial feedback for software maintenance and evolution, but manually extracting useful reviews from vast volumes is time-consuming and challenging. This study investigates the effectiveness of six Naïve Bayes variants for automatically filtering useful app reviews. We evaluated these variants on datasets from five popular apps, comparing their performance in terms of accuracy, precision, recall, F-measure, and processing time. Our results show that Expectation Maximization-Multinomial Naïve Bayes with Laplace smoothing performed best overall, achieving up to 89.2% accuracy and 0.89 F-measure. Complement Naïve Bayes with Laplace smoothing demonstrated particular effectiveness for imbalanced datasets. Generally, incorporating Laplace smoothing and Expectation Maximization improved performance, albeit with increased processing time. This study also examined the impact of data imbalance on classification performance. Our findings suggest that these advanced Naïve Bayes variants hold promise for filtering useful app reviews, especially when dealing with limited labeled data or imbalanced datasets. This research contributes to the body of evidence around app review mining and provides insights for enhancing software maintenance and evolution processes.

Atmospheric pressure plasma jet deposition of TiO2 layer on alumina/epoxy to improve electrical properties

Abstract In gas-insulated lines, basin-insulators can accumulate charge under non-uniform electric fields, distorting the field distribution and potentially causing surface flashover, which threatens the stability of power systems. In this study, Atmospheric Pressure Plasma Jet (APPJ) technology was used to deposit TiO2 on the surface of alumina/epoxy (Al2O3/EP) composites. The impact of deposition of TiO2 layer on the surface morphology and chemical composition of Al2O3/EP was studied using testing methods such as Scanning Electron Microscope, X-ray photoelectron spectroscopy, Fourier Transform Infrared Spectrometer, and Energy Dispersive Spectrometer. It was found that APPJ creates a dense, rough Ti-O layer on the Al2O3/EP surface, which bonds tightly with the substrate. The efficacy of APPJ was found to depend on processing time, with optimal results observed at 3 min, DC and AC flashover voltages increased by 29.6% and 15.7%, respectively. TiO2 layer enhances the conductivity of the resin and shallows trap levels. Through the synergistic effects of various factors, surface charges are efficiently dissipated and evenly distributed. This study not only reveals the physicochemical process of TiO2 deposition via APPJ but also integrates surface characteristics with electrical performance. The findings offer a new strategy to enhance surface flashover voltage and ensure equipment safety.

Privacy-Preserving Data Analytics in Internet of Medical Things

The healthcare sector has changed dramatically in recent years due to depending more and more on big data to improve patient care, enhance or improve operational effectiveness, and forward medical research. Protecting patient privacy in the era of digital health records is a major challenge, as there could be a chance of privacy leakage during the process of collecting patient data. To overcome this issue, we propose a secure, privacy-preserving scheme for healthcare data to ensure maximum privacy of an individual while also maintaining their utility and allowing for the performance of queries based on sensitive attributes under differential privacy. We implemented differential privacy on two publicly available healthcare datasets, the Breast Cancer Prediction Dataset and the Nursing Home COVID-19 Dataset. Moreover, we examined the impact of varying privacy parameter (ε) values on both the privacy and utility of the data. A significant part of this study involved the selection of ε, which determines the degree of privacy protection. We also conducted a computational time comparison by performing multiple complex queries on these datasets to analyse the computational overhead introduced by differential privacy. The outcomes demonstrate that, despite a slight increase in query processing time, it remains within reasonable bounds, ensuring the practicality of differential privacy for real-time applications.

Effect of the estimation result of the degree of polarization of target light on clear imaging

Previous underwater imaging methods have not developed a clear idea of estimating the degree of polarization of target light (Pobj). To address this issue, this Letter answers the question of how the estimation result of Pobj affects clear imaging. First, the theoretical derivation states that Pobj is simply a scale modulation factor of the imaging result. Second, experiments are conducted for validation, and results conform well to the derivation. Hence, the effect of the estimated Pobj on clear imaging is obtained. This parameter only influences brightness rather than contrast but may cause noise amplification as well as the unfavorable result of negative pixels. Therefore, no precise estimation is needed; pick the value near the ends of the definition domain directly and take the absolute value. Based on these, a new imaging formula is proposed, enabling the processing time to fulfill the actual dynamic imaging requirements. As far as we are concerned, the attained prior knowledge and formula could provide strong assistance for underwater polarization imaging.

Enhanced Curvature-Based Fabric Defect Detection: A Experimental Study with Gabor Transform and Deep Learning

Quality control at every stage of production in the textile industry is essential for maintaining competitiveness in the global market. Manual fabric defect inspections are often characterized by low precision and high time costs, in contrast to intelligent anomaly detection systems implemented in the early stages of fabric production. To achieve successful automated fabric defect identification, significant challenges must be addressed, including accurate detection, classification, and decision-making processes. Traditionally, fabric defect classification has relied on inefficient and labor-intensive human visual inspection, particularly as the variety of fabric defects continues to increase. Despite the global chip crisis and its adverse effects on supply chains, electronic hardware costs for quality control systems have become more affordable. This presents a notable advantage, as vision systems can now be easily developed with the use of high-resolution, advanced cameras. In this study, we propose a discrete curvature algorithm, integrated with the Gabor transform, which demonstrates significant success in near real-time defect classification. The primary contribution of this work is the development of a modified curvature algorithm that achieves high classification performance without the need for training. This method is particularly efficient due to its low data storage requirements and minimal processing time, making it ideal for real-time applications. Furthermore, we implemented and evaluated several other methods from the literature, including Gabor and Convolutional Neural Networks (CNNs), within a unified coding framework. Each defect type was analyzed individually, with results indicating that the proposed algorithm exhibits comparable success and robust performance relative to deep learning-based approaches.

Technological line for the production of sunflower cake

BACKGROUND: The results of the study conducted to increase the shelf life of sunflower cake by justifying the optimal operating parameters of the ozonizer in the line of the sunflower cake production are presented. Studies of the processing of sunflower cake with an ozone-air mixture at optimal operating parameters, such as its concentration, the duration of its processing and the intensity of the high-frequency field, have been conducted. AIM: Increasing of the long-term storage of protein feed by improving the basic design parameters of the device for its production. METHODS: Resulting from the conducted theoretical research, the technological line for the protein feed production was developed according to the patent for the invention of the Russian Federation No. 2706188. To determine the optimal operating parameters of the device in the sunflower cake production line, which affect its shelf life, the design of experiment theory was used. RESULTS: A regression equation in coded values, characterizing the effect of the concentration of the ozone-air mixture, the duration of the sunflower cake treatment and the intensity of the high-frequency field on the sunflower cake shelf life, is obtained. The analysis of the obtained mathematical model of the processing process is performed. CONCLUSION: Experimental studies of the developed line for the production of sunflower cake with quality control of its treatment with an ozone-air mixture have shown that the sunflower cake shelf life depends on the ozone-air mixture concentration and the duration of treatment of the sunflower cake, as well as on the intensity of the high-frequency field. At the same time, with a decrease in the ozone-air mixture concentration, the duration of treatment of the sunflower cake and the intensity of the high-frequency field during the treatment increase. In the area of low concentration, increased duration up to 120 s and intensity up to 15250 V/m, the shelf life of the cake increases. It is reasonable to process sunflower cake with the ozone-air mixture to increase its shelf life from 3 to 5.45 months according to the GOST 80-96 using the developed line (patent RF 2706188) at a concentration of 23.03 mg/m3, processing time of 134 s and a high-frequency field strength of 17374 V/m.

An efficient deep reinforcement learning based task scheduler in cloud-fog environment

Efficient task scheduling in cloud and fog computing environments remains a significant challenge due to the diverse nature and critical processing requirements of tasks originating from heterogeneous devices. Traditional scheduling methods often struggle with high latency and inadequate processing times, especially in applications demanding strict computational efficiency. To address these challenges, this paper proposes an advanced fog-cloud integration approach utilizing a deep reinforcement learning-based task scheduler, DRLMOTS (Deep Reinforcement Learning based Multi Objective Task Scheduler in Cloud Fog Environment). This novel scheduler intelligently evaluates task characteristics, such as length and processing capacity, to dynamically allocate computation to either fog nodes or cloud resources. The methodology leverages a Deep Q-Learning Network model and includes extensive simulations using both randomized workloads and real-world Google Jobs Workloads. Comparative analysis demonstrates that DRLMOTS significantly outperforms existing baseline algorithms such as CNN, LSTM, and GGCN, achieving a substantial reduction in makespan by up to 26.80%, 18.84, and 13.83% and decreasing energy consumption by up to 39.60%, 30.29%, and 27.11%. Additionally, the proposed scheduler enhances fault tolerance, showcasing improvements of up to 221.89%, 17.05%, and 11.05% over conventional methods. These results validate the efficiency and robustness of DRLMOTS in optimizing task scheduling in fog-cloud environments.

Coupling of saccade plans to endogenous attention during urgent choices

The neural mechanisms that willfully direct attention to specific locations in space are closely related to those for generating targeting eye movements (saccades). However, the degree to which the voluntary deployment of attention to a location necessarily activates a corresponding saccade plan remains unclear. One problem is that attention and saccades are both automatically driven by salient sensory events; another is that the underlying processes unfold within tens of milliseconds only. Here, we use an urgent task design to resolve the evolution of a visuomotor choice on a moment-by-moment basis while independently controlling the endogenous (goal-driven) and exogenous (salience-driven) contributions to performance. Human participants saw a peripheral cue and, depending on its color, either looked at it (prosaccade) or looked at a diametrically opposite, uninformative non-cue (antisaccade). By varying the luminance of the stimuli, the exogenous contributions could be cleanly dissociated from the endogenous process guiding the choice over time. According to the measured time courses, generating a correct antisaccade requires about 30 ms more processing time than generating a correct prosaccade based on the same perceptual signal. The results indicate that saccade plans elaborated during fixation are biased toward the location where attention is endogenously deployed, but the coupling is weak and can be willfully overridden very rapidly.

MIPPIS: protein–protein interaction site prediction network with multi-information fusion

BackgroundThe prediction of protein–protein interaction sites plays a crucial role in biochemical processes. Investigating the interaction between viruses and receptor proteins through biological techniques aids in understanding disease mechanisms and guides the development of corresponding drugs. While various methods have been proposed in the past, they often suffer from drawbacks such as long processing times, high costs, and low accuracy.ResultsAddressing these challenges, we propose a novel protein–protein interaction site prediction network based on multi-information fusion. In our approach, the initial amino acid features are depicted by the position-specific scoring matrix, hidden Markov model, dictionary of protein secondary structure, and one-hot encoding. Simultaneously, we adopt a multi-channel approach to extract deep-level amino acids features from different perspectives. The graph convolutional network channel effectively extracts spatial structural information. The bidirectional long short-term memory channel treats the amino acid sequence as natural language, capturing the protein’s primary structure information. The ProtT5 protein large language model channel outputs a more comprehensive amino acid embedding representation, providing a robust complement to the two aforementioned channels. Finally, the obtained amino acid features are fed into the prediction layer for the final prediction.ConclusionCompared with six protein structure-based methods and six protein sequence-based methods, our model achieves optimal performance across evaluation metrics, including accuracy, precision, F1, Matthews correlation coefficient, and area under the precision recall curve, which demonstrates the superiority of our model.

Comprehensive Survey of Image Steganography Systems based on FPGA Implementation

Steganography is the art of concealed communication. The basic idea of steganography is to hide secret data inside a cover media in an undetectable manner. Cover media and Secret data can be any popular digital media such as image, video, audio, or text. Various types of image steganography methods have been proposed according to the cover image domain, i.e. spatial domain and transform domain. The spatial domain methods are easy and simple while the transform domain methods are more complex. However, the transform domain methods are more robust against image processing operations. Besides, they are more secure and less detectable in an unsecured channel. The recent steganography methods are based on sophisticated algorithms that include several computational time-consuming tasks, and hence they are unable to embed and extract the hidden data in real-time. Therefore, the trend is to implement the steganography methods in the hardware to speed up their processing time and so improves the efficacy of steganography techniques. FPGA is used in various fields in the modern era due to lower development cost, flexibility, and reconfigurability. Most of the published steganography techniques carried out on FPGA are based on the spatial domain. In this paper, it has comprehensively studied and reviewed various existing implementations of image steganography on FPGAs. This includes studying their general operations, requirements, and performance evaluations. This review would assist researchers in finding the research gaps in FPGA implementation of steganography methods for real-time applications.

TopoSense: agent driven topological graph extraction from remote sensing image

Abstract. Automatic topological graph extraction is critical for intelligent remote sensing image interpretation and cartographic representation. However, existing approaches neither adopt segmentation-based post-processing nor directly predict the graph, thereby suffering from limited scalability and poor adaptability to complex spatial structures. To address these issues, we introduce TopoSense, an innovative framework for extracting topological graphs from remote sensing images through an agent-driven approach. By employing a novel combination of reinforcement learning and neural network architectures, TopoSense autonomously navigates through pixel-level data, efficiently constructing topological representations. It not only enhances the accuracy of spatial feature detection, but also significantly reduces processing time. Experiments on the TOP-BOUNDARY and REALSCENE demonstrate its superiority in capturing intricate spatial relationships compared to traditional methods.

Performance enhancement in blockchain based IoT data sharing using lightweight consensus algorithm

The proliferation of Internet of Things (IoT) devices generates vast amounts of data, traditionally stored, processed, and analyzed using centralized systems, making them susceptible to attacks. Blockchain offers a solution by storing and securing IoT data in a distributed manner. However, the low performance and poor scalability of blockchain technology pose significant challenges for its application in IoT networks. The primary obstacle is the distributed consensus protocol, while ensuring data transparency, integrity, and immutability in a decentralized and untrusted circumstances which often compromises scalability. To address this issue, this paper introduces the use of the Delegated Proof of Stake (DPoS) consensus algorithm and sharding techniques to enhance scalability in blockchain-based IoT networks. Experimental results indicate that system throughput increases synchronously with the test load. Our findings reveal a tradeoff between throughput, latency, and up-downstream time on the Inter Planetary File System (IPFS). Given the critical importance of latency and throughput in IoT networks, the results demonstrate that DPoS offers high throughput, parallel processing, and robust security while efficiently scaling the network. Furthermore, at a test load of 500 Transactions Per Second (TPS), the system achieves a maximum throughput of approximately 11.094 ms. However, when the test load exceeds 2000 TPS, the total processing time for transactions extends to 11.205 ms. This method is particularly suitable for constrained IoT networks. Compared to previous edge computing-based approaches, our scheme demonstrates superior throughput performance.

Energy-efficient time and cost constraint scheduling algorithm using improved multi-objective differential evolution in fog computing

The recent surge in Internet of Things (IoT) applications and smart devices has led to a substantial rise in the data generation. One of the major issues involved is to meet strict quality of service (QoS) requirements for computing these applications in terms of execution time, cost and in an energy-efficient manner. To extract useful information, fast processing and analysis of data is needed. Consequently, moving all the data to centralized cloud data centers would lead to high processing times, increased cost and energy consumption and more bandwidth usage; thus, processing of applications with strict latency requirements becomes challenging. The addition of fog layer between cloud and IoT devices has provided promising solutions to such issues. However, efficient employment of computing resources in the hybrid infrastructure of fog and cloud nodes is of great significance and demands an optimal scheduling strategy. Toward this direction, a novel Pareto-based algorithm in fog computing, namely energy-efficient time and cost (ETC) constraint scheduling algorithm, is introduced in this paper for scheduling workflow applications. ETC attempts to optimize monetary cost along with time and energy objectives. Improved multi-objective differential evolution (I-MODE) meta-heuristic is introduced and incorporated with deadline-aware stepwise frequency scaling approach that is based on our previously proposed energy makespan multi-objective optimization (EM-MOO) algorithm. Synthetic and real-world application workflows are used to conduct evaluation of the proposed work with existing well-known algorithms from the literature. The experimental results for synthetic workflows reveal that the proposed algorithm lessens energy utilization by 14–21%, execution time by almost 25% and cost consumption by 22–27%, while for real-world application workflows, energy consumption is reduced by 12–24%, execution time by 14–16% and cost consumption by 23–29%.

Development of a Key Method for the Optimization of Port Vessel Detection Based on an Improved Multi-Structural Morphology Approach

All over the world, many ports have implemented surveillance camera systems to monitor the vessels and activities around them. These types of systems are not very effective in accurately detecting activities around the port due to background noise and congestion interference at the sea surface. This is why it is difficult to accurately detect vessels, especially smaller vessels, when it turns dark. It is known that some vessels do not comply with maritime rules, particularly in port and safety zones; these must be detected to avoid incidents. For these reasons, in this study, we propose and develop an improved multi-structural morphology (IMSM) approach to eliminate all of this noise and interference so that vessels can be accurately detected in real time. With this new approach, the target vessel is separated from the sea surface background through the weighted morphological filtering of several datasets of structural components. Then, neighborhood-based adaptive fast median filtering is used to filter out impulse noise. Finally, a characteristic morphological model of the target vessel is established using the connected domain; this allows the sea surface congestion to be eliminated and the movement of vessels to be detected in real time. Multiple tests are carried out on a small and discrete area of moving vessels. The results from several collected datasets show that the proposed approach can effectively eliminate background noise and congestion interference in video monitoring. The detection accuracy rate and the processing time are improved by approximately 3.91% and 1.14 s, respectively.

Estimating Stellar Atmospheric Parameters and [α/Fe] for LAMOST O-M-type Stars Using a Spectral Emulator

In this paper, we developed a spectral emulator based on the Mapping Nearby Galaxies at Apache Point Observatory Stellar Library (MaStar) and a grouping optimization strategy to estimate effective temperature (T eff), surface gravity (log g), metallicity ([Fe/H]), and the abundance of alpha elements with respect to iron ([α/Fe]) for O-M-type stars within the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) low-resolution spectra. The primary aim is to use a rapid spectral-fitting method, specifically the spectral emulator with the grouping optimization strategy, to create a comprehensive catalog for stars of all types within LAMOST, addressing the shortcomings in parameter estimations for both cold and hot stars present in the official LAMOST AFGKM-type catalog. This effort is part of our series of studies dedicated to establishing an empirical spectral library for LAMOST. Experimental results demonstrate that our method is effectively applicable to parameter prediction for LAMOST, with the single-machine processing time within 70 hr. We observed that the internal error dispersions for T eff, log g, [Fe/H], and [α/Fe] across different spectral types lie within the ranges of 15–594 K, 0.03–0.27 dex, 0.02–0.10 dex, and 0.01–0.04 dex, respectively, indicating a good consistency. A comparative analysis with external data highlighted deficiencies in the official LAMOST catalog and issues with MaStar parameters, as well as potential limitations of our method in processing spectra with strong emission lines and bad pixels. The derived atmospheric parameters as a part of this work are available via doi:10.12149/101402.

The Effect of Lean Manufacturing on Production / Operation for the Small and Medium Enterprise in Malaysia

At the start of 2022, the government announced its focus on supporting the development of small and medium enterprises (SMEs) to improve living standards following the COVID-19 pandemic. SMEs contribute significantly to the national economy, with the clothing industry playing a major role. This research applies lean manufacturing techniques to enhance the performance of SMEs in the clothing sector, specifically in T-shirt printing. Therefore the first objective was to collect, analyze processing times and recording of all issues encountered during production. The second objective involved using value stream mapping (VSM) to analyze the T-shirt printing process from raw material supply to finished product packaging. We also employed a why-why analysis to pinpoint the underlying causes of inefficiencies and devise specific solutions. The last objective focused on process improvement using VSM, along with the Kanban system, 5S and Make to Stock (MTS) to optimize block preparation and T-shirt printing. We selected Flexsilk Printshop as the case study. The findings showed a 66.59 % reduction in waiting time, a 30.77 % improvement in lead time, an increase in overall equipment effectiveness (OEE) from 93.3 to 97.45 % and a 50.41 % increase in efficiency.

Ultraviolet line structuring of large solar wafers

AbstractUltrashort‐pulsed lasers are ideal for material processing by welding, marking, and cutting. Scanning systems are crucial for large substrates. We introduce a novel UV line structuring method using deformable mirror technology and freeform optics. This approach aims to improve processing times on large substrates, especially for M12 solar wafer processing. By using a freeform lens and a Zwobbel deformable mirror, we achieve extended processing fields and high scan frequencies. The paper explores optical designs, their robust optomechanical implementation for prototyping, and setup characterization.

Artificial Intelligence Chatbots in Action: Optimizing Benefits Enrollment in Public Administration

This article examines the implementation of artificial intelligence-powered chatbots in public sector benefits enrollment processes, focusing on their potential to streamline operations, enhance user experience, and reduce administrative burdens. Through a comprehensive analysis of case studies in healthcare and social security benefits programs, we demonstrate that chatbots can significantly improve efficiency, accuracy, and accessibility in enrollment procedures. Our findings indicate a 30-40% reduction in processing times and a marked increase in application completion rates. However, challenges related to data privacy, system integration, and user acceptance persist. This article contributes to the growing body of literature on digital transformation in public administration by providing empirical evidence of chatbots' effectiveness in benefits enrollment. We conclude that while chatbots offer promising solutions to longstanding issues in public sector service delivery, their successful implementation requires careful consideration of technical, ethical, and user-centric factors. Our study has important implications for policymakers and public administrators seeking to leverage AI technologies to enhance public service efficiency and accessibility