https://www.jdcs.ir/ Computing and distributed systems en daily 1 Fri, 20 Feb 2026 00:00:00 +0330 Fri, 20 Feb 2026 00:00:00 +0330 https://www.jdcs.ir/article_229635.html Given the increasing number of jobs and job seekers, it seems necessary to have a system that can provide suitable jobs to job seekers on the web. The best way to do this is to use recommender systems. Recommender systems can provide users with suggestions that are of interest to them. In this study, we present a new method to improve recommender systems in the field of job suggestions to users. The working method is that we first collect the records of individuals' jobs, then classify this data using data mining algorithms and according to each person's interest. Then, we will provide this data to the user using a recommender system based on collaborative filtering. The results of the evaluation of the proposed method indicate the high performance of this proposed system. In such a way that the proposed method was able to achieve an accuracy rate of 92% and a recall rate of 96%, and in general, this system can provide up to 90% of recommendations to users that can be of interest to the target user to a high percentage. https://www.jdcs.ir/article_237900.html In today's world, due to the wide range and diversity of resources and products, meeting the needs of systems is a complex and difficult matter. Therefore, a chain of several factors needs to be formed to meet the process of meeting the needs of a system. Such a chain is called a supply chain. It is obvious that supply chain management and optimization can meet the needs optimally. So far, numerous solutions, methods, and models have been presented for optimizing supply chains. But in most of them, supply chains are considered as one-stage. That is, they assume the supply chain to be fully integrated or centrally managed, and as a result, potential fluctuations in the production processes of suppliers and the needs of customers are ignored. Fundamental changes have occurred in contemporary agricultural, manufacturing, industrial, and commercial environments. These changes have affected supply chains and caused them to face many challenges. One of these challenges is that in many cases, supply chains cannot meet the needs of customers in one stage. Rather, they have to meet orders in several stages. That is, today, in some systems, due to problems in meeting needs, they are forced to use multi-stage supply chains. In such supply chains, the needs of the final stage are met by regular and sequential activities of several stages in a row. Therefore, it is good to conduct research on the management and optimization of multi-stage supply chains. https://www.jdcs.ir/article_236368.html Distributed systems, as the backbone of modern information technology, provide the critical infrastructure for cloud services, the Internet of Things, digital financial networks, and large-scale computing. Despite their central role, such systems continually face challenges such as hardware and software failures, communication latency, message asynchrony, and the inherent dynamism of the execution environment. These challenges intensify the need for robust solutions that ensure fault tolerance, data consistency, and functional integrity.Adopting a structured analytical review approach, the present study systematically examines failure models, fault-detection mechanisms, and consensus algorithms in distributed systems. The findings indicate that the design of adaptive and intelligent failure detectors plays a pivotal role in enhancing system stability and reliability. Moreover, the results show that integrating adaptive detectors with lightweight consensus algorithms such as Raft and PBFT provides an effective pathway toward achieving resilient distributed systems in dynamic environments.In addition, the use of machine learning algorithms for intelligent fault prediction and detection, as well as the integration of blockchain technology with classical consensus mechanisms, is proposed as a set of emerging research directions aimed at improving security, efficiency, and scalability. The outcomes of this research can serve as both a theoretical and practical foundation for the design and implementation of distributed infrastructures that exhibit high levels of resilience, self-regulation, and fault tolerance while maintaining effective and intelligent performance in the face of environmental fluctuations. https://www.jdcs.ir/article_240059.html The integration of the Internet of Things (IoT) into the healthcare ecosystem has precipitated a paradigm shift in patient treatment trajectories. By facilitating rapid and highly reliable service delivery, IoT has fundamentally revamped the healthcare landscape. Current research indicates that the proliferation of sensor-based medical devices is continuously enhancing the reliability and efficiency of therapeutic processes. Driven by rapid advancements in communication technologies, the Internet of Medical Things (IoMT) has assumed a pivotal role in the digital transformation of the industry. However, this evolution is accompanied by multifaceted challenges, including the need for lightweight hardware, ensuring data and model explainability, and addressing critical security and privacy concerns. To fully leverage IoMT’s potential, it is imperative to focus on optimizing algorithms for minimal power consumption, developing novel explainable AI models, and ensuring resilience against security threats. This paper presents a comprehensive review of IoMT, detailing the expansion of smart healthcare systems, diverse applications, and core components. Furthermore, it provides a critical analysis of related works to highlight existing gaps and future directions in medical IoT. https://www.jdcs.ir/article_240035.html Smart learning is a common feature in educational environments that arises from the emergence of new technologies and is used in many sectors, such as education, in the learning process and to improve the quality of education. The phenomenon of the Internet of Things today plays an important role in many fields and has made the surrounding environment more innovative and responsive, which has improved life. This technology can be considered as a large network with various types of connected objects that are able to communicate with each other and exchange information, regardless of whether they belong to the same group or not. Creating a network consisting of interconnected devices allows the user to manage all connected devices more effectively. This review examines the integration and impacts of the Internet of Things in education, highlighting its importance in transforming traditional teaching and learning techniques, and examines the early applications and historical growth of the Internet of Things, its development and milestones in its adoption. It shows how the Internet of Things can create more personalized learning paths, enhance student engagement, and foster a connected and sustainable learning environment through the use of modern technology. In fact, the aim of this article is to present the role of the Internet of Things in achieving smart environments in various contexts, with special attention paid to the concept of smart schools and smart education. https://www.jdcs.ir/article_240058.html Traditionally, static models are used to model credit bank patterns, but economic factors are independent of political fluctuations; as the political atmosphere changes, the economic environment also changes with it. This is especially evident in Iran after the 2008-2016 US sanctions, as it is highly likely that they were unable to repay their debts (i.e. became bad customers). It is necessary to create dynamic models that incorporate various political-economic factors, which can be combined to propose hybrid models for credit assessment relative to bank classification based on classification algorithms and proposed algorithms. The model can have two stages, the first stage of which is data preprocessing (to eliminate data defects) resulting from the feature selection function, that is, selecting the appropriate features from among the existing features in the feature set. The second stage of the proposed system is the result of the work and the main output of the proposal, namely the classification of customers, who are divided into two categories, namely those who are eligible to borrow and those who have the conditions to qualify for borrowing, where appropriate classification techniques can be used in this part. In fact, the main objective of this review is to propose models for credit prediction and assessment in terms of bank classification based on machine learning and developmental algorithms. https://www.jdcs.ir/article_240057.html Creditworthiness is one of the most vital and important tasks in the modern banking industry, which ensures the non-performing and sustainability of credit financial institutions with an accurate prediction of the credit status of loan applicants.The main goal of this research is to improve credit assessment for small companies, which is an innovative measure in this field.In fact, to achieve this goal, comprehensive datasets including a variety of financial information, micro-company information, public credit information, and third-party personal access to information are required so that unbalanced sample processing techniques can be used to achieve fairer samples. In this regard, a crow algorithm is used to solve the feature selection problem and in the second step, the selected features are provided as input to a convolutional neural network for final classification. The proposed model was implemented and evaluated on a real-time validated dataset from the Kaggle platform, which included real banking information. K-Fold cross-validation method was used to accurately measure the model performance. The experimental results show that this hybrid model has achieved significantly higher accuracy compared to traditional classification methods. Finally, this research showed that the proposed model has high potential for use as a decision support system in banks and financial institutions for credit assessment. https://www.jdcs.ir/article_229636.html The purpose of this research is to investigate the impact of comprehensive quality managementon the performance of the organization through human resource competence and capabilityThis research is an innovation in the industrial production company of Ava-MedikIt is an applied research in terms of purpose and in terms of nature and methodIt is a descriptive-survey research. Statistical research population480 experts of the industrial production company of Ava-Medratare peopleAccording to Cochran's formula, the sample size is 213 peoplehas been obtained. From the questionnaire to collect dataResearch has been used. For the validity of the case questionnaire researchThe use has been standard, but to ensure validityAlso localization, this questionnaire is available to a number of professorsand the CVR (content validity) of the questionnaires was takenFor the reliability of the research, the value of Cronbach's alpha is 0.772Therefore, the validity and reliability of the questionnaires have been confirmed.Analysis of the results obtained in this research using softStatistical softwares are performed at two descriptive and inferential levelsis At the descriptive level of statistics such as frequency and percentageFrequency is used. At the level of inferential statistics for the testResearch hypotheses from structural equations and Lisrel softwarehas been usedThe importance of comprehensive quality on innovation capabilityompetence of human resourcesYes. The ability to innovate has an impact on the performance of the company. https://www.jdcs.ir/article_241021.html The load balancing challenge in cloud computing is very important for the effective management of cloud resources and it requires distributing the incoming network traffic or computational workload among multiple servers in a way that no server is overloaded, hence improving resource utilization, increasing throughput and reducing the response time of load balancing is crucial for distributed systems to achieve high availability and fault tolerance. In this paper, a task scheduling model in grid networks based on the Lion on the Optimization is presented. The aim of this paper is to present an improved Lion on the Optimization metaheuristic approach to solve load balancing problems in the cloud. In fact, the proposed method addresses the challenge of a single-machine scheduling problem in which the machine must be maintained after a fixed periodic interval. The aim is to minimize the total absolute deviation of the completion times for the problem under consideration. Hence, to solve this problem, a new metaheuristic algorithm, namely Lion on the Optimization, was used in the proposed method, and the comparative results showed the absolute superiority of our proposed algorithm over the base paper. https://www.jdcs.ir/article_241022.html Since the introduction of so-called non-fungible tokens, a digital artwork can be worth several million dollars. These digital tokens are unique, representing ownership of certain unique digital objects. In this paper, we propose an approach to predict the sales of non-fungible token products based on neural networks, which addresses the equivalence and prediction of pricing between real and virtual effects and predict their sales. In the proposed method, an application of hedonic pricing modeling to the spatial properties of virtual assets is demonstrated, which is the first of its kind. Previous studies have only examined the volume-return or volume-volatility relationship of fungible tokens, such as Bitcoin and other “regular” cryptocurrencies. However, the results of this study contribute to a better understanding of the volume-return and volume-volatility relationships of non-fungible tokens, namely Theta, Tezos, and Enjincoin, which are built on alternative blockchains and are gaining popularity among the non-fungible token communities of the Ethereum protocol. In this study, neural networks were used for classification and good results were obtained in terms of accuracy or precision for predicting the sales of non-fungible token products based on this technique. Also, comparing the proposed method and other similar methods for predicting the price of non-fungible tokens in different evaluation criteria resulted in more optimal and appropriate results. https://www.jdcs.ir/article_241040.html Vehicular Ad-hoc Networks are networks that are created by combining moving vehicles and related infrastructures and have challenges, the most important of which are routing and power reduction due to the dynamic nature of this network, which is due to the increase in the number of nodes. In this research, clustering methods such as K-Means and evolutionary algorithms are used to improve routing in these networks. In fact, the K-Means algorithm divides the network into smaller parts (several smaller clusters) and a cluster head is generated in each of the clusters, which is used to transfer information between vehicles inside and outside a cluster, and choosing a suitable and reliable cluster head is of great importance. Therefore, the Gray Wolf algorithm is used to choose the best and most suitable cluster head for each of the clusters. In the simulation of the proposed method, various experiments were conducted to investigate the effect of the number of data transmission clusters, packet delivery rate, message sending delay, and message number on the transmission performance. The results showed that by increasing the number of vehicles in the proposed method, fewer clusters are required, the path of each message and the stopping time at each step are stored, and the proposed method has better performance than the basic method. https://www.jdcs.ir/article_241436.html Automatic extraction of software requirements from natural language texts remains a central challenge in Requirements Engineering due to ambiguity, polysemy, and the heterogeneity of information sources. In recent years, approaches based on Natural Language Processing (NLP) have provided more controllable and structured outputs; however, they face limitations when dealing with implicit expressions, complex sentence structures, and unstructured data. In contrast, Large Language Models (LLMs), with their semantic understanding and reasoning capabilities, demonstrate strong potential for processing complex textual content and conversational data. Nevertheless, output instability, prompt sensitivity, and the risk of generating inaccurate or fabricated content make their direct application in engineering contexts challenging.This study adopts an analytical–comparative approach to examine a reference framework based on a large language model and compares it with three representative approaches for requirements extraction from documents and user feedback. The findings indicate that although LLM-based approaches offer advantages in handling unstructured data, achieving reliable outputs requires standardization mechanisms and robust quality control strategies. Accordingly, the study emphasizes the necessity of developing hybrid approaches that integrate semantic intelligence with structured control mechanisms.