HighTech and Innovation Journal
Not a member yet
317 research outputs found
Sort by
Deep Learning: A Study of Pattern Recognition for Personalized Clothing
Full text linkObjectives: This article aims to enhance the efficiency of clothing recognition and retrieval by implementing deep learning algorithms for personalized clothing pattern recognition. Methods: Based on the you only look once version 4 (YOLOv4) algorithm in deep learning, the CSPDarknet-53 in the original algorithm was replaced by GhostNet, and the original Leaky ReLU activation function was replaced by FMish. Then, an improved YOLOv4 algorithm was obtained. Experiments were carried out on the personalized clothing pattern set, the Fashion Mnist dataset, and the DeepFashion dataset to compare and analyze different algorithms. Findings: When replacing CSPDarknet-53 with GhostNet and the Leaky ReLU activation function with FMish, the optimized YOLOv4 algorithm performed significantly better, verifying the reliability of the YOLOv4 improvement. The optimized algorithm achieved an F1 value of 94.22% and a mAP of 95.41% on different datasets, and 39.51% and 49.56% on the DeepFashion dataset, respectively, outperforming other deep learning methods such as the faster-recurrent convolutional neural network. Furthermore, the floating-point operations per second of the optimized YOLOv4 algorithm were 8.72 G, showing a reduction of 49.71% compared to the traditional algorithm. This suggested that it had low complexity and calculation amounts. Novelty: The optimized YOLOv4 algorithm performs excellently in recognizing personalized clothing patterns, which can provide a new and reliable approach for recognition and retrieval in the field of clothing. Doi: 10.28991/HIJ-2023-04-03-03 Full Text: PD
Big Data Analysis using Elasticsearch and Kibana: A Rating Correlation to Sustainable Sales of Electronic Goods
Full text linkBig data collection involves enormous amounts of raw data. To boost the sustainability of corporate value and support business intelligence and decision-making systems, in-depth data analysis is necessary. The data storage, analysis, and visualization methods, as well as the discovery of patterns and linkages, all depend on extensive data analysis. This study aims to process datasets to learn things like how ratings impact market sales transactions and how much of an impact factor connected to consumers and items have on ratings. Elasticsearch and Kibana were used for the dataset processing. This study evaluated traits related to the test parameters using a variety of test procedures. The product is scored as a representation of the product types involved in the sales transaction, and the name is assessed as a reflection of the customer. Kibana and Elasticsearch, a full-text search engine, were used in this work to do extensive data analysis on data sets. It is a visualization tool that is employed in a controlled environment to evaluate how ratings impact market exchanges for electronic goods, and it offers suggestions. The study found a substantial relationship between electronic product sales on the Amazon marketplace from 2012 to 2018. It suggested the importance of buyer constituents as users and how different product categories relate to ratings in business transactions. Doi: 10.28991/HIJ-2023-04-03-09 Full Text: PD
Enhancing Trustworthiness and Interoperability of Electronic Voting Systems through Blockchain Bridges
Full text linkDecentralized applications leveraging blockchain technology are gaining widespread adoption within the decentralized applications ecosystem. Interoperability, a fundamental concept facilitating seamless data and processing power exchange across diverse blockchain networks, is paramount in this context. The primary objective of this paper is to explore the transformative potential of "blockchain bridges" in facilitating secure and transparent electronic voting processes across multiple blockchain networks. The study employs a comprehensive analysis of various approaches, including atomic exchanges, sidechains, cross-chain bridges, token wrappers, and interledger protocols. The selection of a specific method is guided by the unique requirements and privacy considerations of the electronic voting use case. The application of two distinct blockchains serves as a practical demonstration, illustrating the principles of blockchain bridges in real-world scenarios. The research reveals that blockchain bridges not only streamline the exchange of data between diverse blockchain networks but also establish a dual decentralization paradigm. This paradigm enables the creation of openly maintained, purpose-specific, decentralized ledgers for electronic voting. The integration of blockchain bridges significantly reduces the risk of fraud, instilling greater confidence in the accuracy of election results. Thus, by presenting a comprehensive array of approaches and emphasizing their practical application, this research contributes to advancing the understanding and implementation of blockchain technology in the critical domain of electronic voting. Doi: 10.28991/HIJ-2023-04-04-04 Full Text: PD
Comprehensive Review of the Impact of Advanced Technology Adoption on Work and Continuous Improvement
Full text linkTechnology advances are changing how companies and their employees do their respective jobs. The intervention of technology has revolutionized the way jobs are conceptualized, discussed, performed, and delivered. This analysis aims to deliver a summary of the impact of the adoption of advanced technology on job characteristics and the impact on job demand and continuous improvement, which can act as a reference for theorists and practitioners to map, research, and analyze the effect of technology on work systems and productivity. By presenting a systematic review of the literature along with several avenues for future research, we hope that this research will contribute to job demand and ongoing research. A total of 30 articles were reviewed from a total generated article database of 335, which were systematically selected from different academic databases between 2001 and 2021. The review signifies the role of technology in influencing work complexities, privacy, workload, workflow interruptions, manual work, role expectations, and developmental opportunities. This study is pivotal in substantiating the influence of technology in work systems, besides furnishing variables and themes for further studies in the area. Doi: 10.28991/HIJ-2023-04-03-014 Full Text: PD
New Technologies and Innovative Solutions in the Development of Multimedia Corpus of Mezen Robinsons Texts
Full text linkObjective: New Technologies and Innovative Solutions in creating a multimedia corpus of texts about the "Mezen Robinsons" aims to preserve the memory of an event that occurred in the 18th century and to study the history of Spitsbergen development. This article presents a multimedia corpus of Russian-language texts about the "Mezen Robinsons" written in 1766–2022. Observations show that the history of the survival of the Mezen hunters on Edge Island in 1743–1749 has repeatedly attracted the attention of specialists from various fields of knowledge: historians, archaeologists, publicists, professional writers, translators, etc. The corpus unites texts, audio, video, and multimedia resources. Methods: continuous sampling was used to collect the material; when analyzing and describing the data, we applied a descriptive method, a biographical method of studying literature, statistical data processing, philological analysis, observation, assessment, and corpus modeling methods. Findings: the methodology and technology of building an independent multimedia corpus, its architecture, and its design are described. Novelty: the multimedia corpus is a contribution to the development of a new approach to studying the subjectology of Russian literature. Practical significance:the findings can become the basis for studying the biographies and creativity of various authors who built their works on the plot of the Mezen industrialists and for further comparison of various interpretations of one event from the history of the development of the Arctic. Doi: 10.28991/HIJ-2023-04-01-07 Full Text: PD
System Architecture for IT Talent Ecosystem Using Service Oriented Approach
Full text linkThe purpose of this research is to propose a System Architecture to facilitate the IT Talent ecosystem using a service-oriented approach. The need for this is important to support digital transformation in the IT Talent ecosystem. Human resources in the IT field are one of the key factors in implementing IT in organizations. However, the availability of IT human resources has not been able to meet the needs and challenges of the organization in synergizing IT and business. Meanwhile, on the other hand, the qualifications of IT human resources do not meet the existing competency standards. In this research, we use a service-oriented system development method. It consists of three stages, such as (1). Analysis and Observation, (2). Analysis from an in-depth interview, and (3) System Architecture Design, which includes Analysis Features of the Systems, Service Analysis and Identification, Specification of Architecture, and Layering. The novelty and findings of this research are a system architecture, which is called a middleware architecture, that can bridge entities in the IT Talent ecosystem to provide and use services to each other for support collaboration. In this study, we proposed a system architecture that acts as middleware to support collaboration and integration in the IT Talent ecosystem. We proposed TALENT-IT, which acts as a service bus mechanism. We used a service-oriented approach to develop this platform. The results of this study are: list of features, list of services, SOA layer, SOA architecture, and monetization feasibility and challenges. Doi: 10.28991/HIJ-2023-04-04-03 Full Text: PD
Research on Customer Relationship Segmentation of Apparel Retail Industry through Data Mining
Full text linkObjectives: This paper aims to segment customers in the apparel retail industry using data mining techniques. Methods: First, a customer segmentation model was constructed, and then the K-means algorithm was used to classify customers based on indicators from the model. The classification effectiveness was enhanced by introducing indicator feature weights. A case study was also conducted. Findings: When the value of K was 4, the K-means algorithm achieved the best classification performance. The improved K-means algorithm outperformed the traditional K-means algorithm in terms of classification effectiveness. The improved K-means algorithm categorized customers into premium customers, important customers, regular customers, and churned customers. Different marketing suggestions were proposed to manufacturers. Novelty: The novelty of this article lies in the introduction of feature weights for indicators, which allows for a distinction between their importance and improves classification effectiveness. Doi: 10.28991/HIJ-2023-04-02-05 Full Text: PD
Seismic Upgradation of RC Beams Strengthened with Externally Bonded Spent Catalyst Based Ferrocement Laminates
Full text linkGlobally, since there are more systems of civil infrastructure, there are also more degraded buildings and structures. If upgrading or strengthening is a practical option, complete replacement is likely to be an escalating financial burden and may be a waste of natural resources. It is necessary to repair or strengthen a number of reinforced concrete buildings and structures in order to boost their load-bearing capabilities or improve their ductility under seismic stress. Additionally, due to changes in service circumstances, a structure might need to be modified to reduce deflections or manage cracking. Strengthening may be preferable to limiting usage, capping applied loads, and regularly inspecting the structure rather than removing the existing structure or part and building a new one. This study aims to examine the flexural, shear, and combined effect of flexural and shear behavior of reinforced concrete (RC) beams strengthened with externally bonded spent catalyst-based ferrocement laminates and compare them to the control beams (unstrengthened) under two-point loading conditions. This study involves researching laminates with various spent catalyst doses, such as 3, 6, 9, and 12%, in an effort to determine the best amounts that will improve the structural performance of ferrocement laminates. Twelve spent catalyst-based ferrocement laminates measuring 500(L) í— 125(B) í— 20 mm (thickness) with 3% volume fraction of meshes each were cast and tested in the lab as part of the preliminary investigation. For repeatability, three laminates per case were employed. Eight numbers of under-reinforced RC beams measuring 75(L) í— 100(B) í— 150(D) mm were cast for the main study; six numbers were strengthened with optimized spent catalyst-based ferrocement laminates bonded with flexible epoxy systems at the tension zone, shear zone, and combination of tension and shear zone. Two of the beams were cast as control specimens. The beams were then evaluated using a Universal Testing Machine (UTM) with a 1000 kN capacity under two-point loading conditions. As a result, the strength, yield load, ultimate load, stiffness, ductility, and related failure modes of all tested beams' flexural and shear performances were examined. According to a preliminary analysis of laminates made of spent catalyst, the dosage of 9% provides good flexural strength in comparison to other doses. In comparison to the strengthened beam, the control beam's initial cracks appeared earlier. In comparison to the control beam, the strengthened beam has an increase in load-carrying capacity of 18% for flexure, 16% for shear, and 30% for the combined impact of flexure and shear. In comparison to the control beam, the deflection of the strengthened beam was decreased by close to 20 to 40% for flexure, 10 to 30% for shear, and 15 to 20% for the combined effects of flexure and shear at the same load level. In relation to control beams, the ductility also improved up to 30% for flexure, 25% for shear, and 25% for the combined impact of flexure and shear. Similar to this, the retrofitted beam is stiffer than the control beam by approximately 40% for flexure, 48% for shear, and 30% for the combined effect of flexure and shear. Theoretical formulation by section analysis is also derived and it gives close agreement with control and strengthened beams. The flexural and shear strengthening of the RC beam retrofitting system is effectively increased by using spent catalyst-based ferrocement laminates. No beam showed signs of premature and brittle failure. According to the test findings, it can be said that spent catalyst-based ferrocement reinforced beams perform better in every way than control beams. Doi: 10.28991/HIJ-2023-04-01-013 Full Text: PD
Finger Vein Template Protection with Directional Bloom Filter
Full text linkBiometrics has become a widely accepted solution for secure user authentication. However, the use of biometric traits raises serious concerns about the protection of personal data and privacy. Traditional biometric systems are vulnerable to attacks due to the storage of original biometric data in the system. Because biometric data cannot be changed once it has been compromised, the use of a biometric system is limited by the security of its template. To protect biometric templates, this paper proposes the use of directional bloom filters as a cancellable biometric approach to transform the biometric data into a non-invertible template for user authentication purposes. Recently, Bloom filter has been used for template protection due to its efficiency with small template size, alignment invariance, and irreversibility. Directional Bloom Filter improves on the original bloom filter. It generates hash vectors with directional subblocks rather than only a single-column subblock in the original bloom filter. Besides, we make use of multiple fingers to generate a biometric template, which is termed multi-instance biometrics. It helps to improve the performance of the method by providing more information through the use of multiple fingers. The proposed method is tested on three public datasets and achieves an equal error rate (EER) as low as 5.28% in the stolen or constant key scenario. Analysis shows that the proposed method meets the four properties of biometric template protection. Doi: 10.28991/HIJ-2023-04-02-013 Full Text: PD
An Improved Differential Evolution Algorithm for Numerical Optimization Problems
Full text linkThe differential evolution algorithm has gained popularity for solving complex optimization problems because of its simplicity and efficiency. However, it has several drawbacks, such as a slow convergence rate, high sensitivity to the values of control parameters, and the ease of getting trapped in local optima. In order to overcome these drawbacks, this paper integrates three novel strategies into the original differential evolution. First, a population improvement strategy based on a multi-level sampling mechanism is used to accelerate convergence and increase the diversity of the population. Second, a new self-adaptive mutation strategy balances the exploration and exploitation abilities of the algorithm by dynamically determining an appropriate value of the mutation parameters; this improves the search ability and helps the algorithm escape from local optima when it gets stuck. Third, a new selection strategy guides the search to avoid local optima. Twelve benchmark functions of different characteristics are used to validate the performance of the proposed algorithm. The experimental results show that the proposed algorithm performs significantly better than the original DE in terms of the ability to locate the global optimum, convergence speed, and scalability. In addition, the proposed algorithm is able to find the global optimal solutions on 8 out of 12 benchmark functions, while 7 other well-established metaheuristic algorithms, namely NBOLDE, ODE, DE, SaDE, JADE, PSO, and GA, can obtain only 6, 2, 1, 1, 1, 1, and 1 functions, respectively. Doi: 10.28991/HIJ-2023-04-02-014 Full Text: PD