International Journal of Computer (IJC - Global Society of Scientific Research and Researchers, GSSRR)
Not a member yet
459 research outputs found
Sort by
Managed File Transfer Solutions: Security and Scalability with AWS Transfer Family
Full text link The study examines Infrastructure as Code for multi-cloud delivery with Terraform and AWS CloudFormation, focusing on conservative cross-cloud abstraction, policy-as-code enforcement, and AI-assisted configuration. Configuration analysis indicates about a 40% reduction in initial setup time and a ~50% decline in recurrent configuration defects. Economic signals show ~15% cost relief for SME tenants and ~30% faster deployment cycles for volatile workloads through pre-validated modules, drift control, and cost guardrails. The paper documents a governance model that maps automated checks to NIST 800-53 control families and integrates plan-time static analysis, secrets detection, and evidence capture. Generative AI is positioned as a CI-embedded assistant that translates natural-language intents into validated templates while remaining policy-, state-, and cost-aware. The contribution consolidates comparative tool behavior, governance placement in the pipeline, and maturity stages for AI-assisted IaC. The material addresses practitioners designing reliable and economical multi-cloud estates and researchers evaluating NL?IaC evaluation workflows
Leveraging Big Data Analytics for Combating Fake News: A Supervised Learning Approach to Identifying Misinformation on Social Media
Full text linkThe rapid rise of social media has transformed how people consume and share information but has also accelerated the spread of misinformation that undermines public trust, public health, and democratic stability. Manual fact-checking and platform moderation often lag behind the speed of misinformation, highlighting the need for scalable, automated solutions. This study develops a supervised machine learning framework supported by Big Data analytics for fake news detection. Using the ISOT Fake News Dataset of 44,898 labeled articles, we implemented a structured pipeline that included text normalization, tokenization, stopword removal, stemming, and TF-IDF vectorization, followed by training four classifiers: Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost). Evaluation was conducted using a stratified 80/20 train-test split with 10-fold cross-validation, applying Accuracy, Precision, Recall, and F1-score as performance metrics. Results show that ensemble models, particularly XGBoost and Random Forest, consistently outperformed LR and SVM, achieving accuracies near 99% with strong precision and recall across both classes. These findings demonstrate the strength of optimized ensemble methods in detecting misinformation and their scalability for real-world application. Beyond model performance, this work proposes a distributed architecture leveraging Apache Spark for real-time deployment, providing a foundation for practical and scalable misinformation detection systems
Infrastructure Deployment Patterns for AI Services on the Claude Platform
Full text linkThis article questions how to select and use infrastructure patterns in the deployment of industrial AI services on Claude due to the increased enterprise demand and broadened capabilities. This study proves relevant since, within a very short span, generative AI work has transformed from novelty experiments to mission-critical business workloads where effectiveness directly depends on architectural decisions ensuring not only low latency but also cost predictability together with the compliance of corporate security requirements. Therefore, the purpose of this paper is to identify recurring architectural patterns and their operational mechanisms that may be used in balancing response time, context size, cost, and reliability when integrating Claude into enterprise IT systems. The scientific novelty of the research is a detailed taxonomy of infrastructural solutions (serverless gateways, containers in isolated VPCs, batch queues) and how their combination begins to meet the very special challenges of Claude—challenges such as context windows up to one million tokens and loads that seem to increase sharply above 25 billion requests per month. The main findings show that the best setup comes from using small APIs with autoscaling and local cache for interactive work; containers with private links and Batch interfaces when dealing with deep analytics and big processing. The quota system and context thinning, breaking up queues by priority and budget, as well as token control and telemetry, are pieces to help give teams a p95 latency promise while staying inside financial and regulatory fences. The article will be helpful to researchers in cloud architectures, enterprise IT system architects, and practitioners deploying AI services based on Claude models
Weapon Classification System using VGG16 and Inception ResNet Models
Full text linkSecurity forces urgently need to implement computerized systems in regard to the increasing number of criminal acts. In the battle against crime, the construction of modern weapon recognizing systems has become crucial. The nature and carefulness of the crime are determined by the type of weapon. In this study, distinct types of weapons classification using deep learning models is presented. The presented approach is developed using the Keras architecture, which is based on the TensorFlow framework, and makes use of the VGGNet and Inception ResNetV2 architecture. The classifier is trained using three classes: knife, gun, and background. The model uses the weapon images from Roboflow. The presented approach outperforms the VGG-16 model (96.25% accuracy) and Inception ResNet-V2 model (97.92% accuracy) in terms of classification accuracy. This study offers a crucial perspective on how well the presented deep learning models handle the challenging issue of weapon classification
Polycrystalline Organic Semiconductors for Low-Power AI-Integrated Devices: Fabrication and Characterization
Full text linkPolycrystalline organic semiconductors constitute a promising class of functional materials, combining architectural flexibility, compatibility with low-temperature processing, and tunable electronic properties. Owing to their ability to form ordered morphologies with defined crystallinity, these materials can be seamlessly integrated into energy-efficient intelligent systems—such as neuromorphic components, sensing nodes, and photo-analytical elements. In the context of rapidly growing interest in edge computing and Internet-of-Things infrastructures, they offer advantages over silicon counterparts, including mechanical robustness, spectral sensitivity, and the capacity for on-site primary signal processing. This study reviews methods for producing polycrystalline films, examines how fabrication parameters influence morphological and electrical structure, and surveys quality-control techniques at both micro- and structural levels. Organic semiconductors thus fulfil a strategic role in the design of adaptive, miniaturized, and self-learning devices operating under unstable conditions, forming the basis for subsequent advances in intelligent organic electronics
Lean MVP Development Via SQL Prototypes: Fast ETL Pipelines, Temporary DWHs, and Accelerated Validation of Product Hypotheses
Full text linkThis paper discusses the use of lightweight SQL prototyping for rapid ETL pipeline construction and MVPs in terms of enabling temporary data warehouses and accelerating product hypothesis validation. It lays out, formalizes, and tests aspects of a Lean Analytical Circuit Building approach based on declarative SQL language such that verifiable metrics may be available to a team without waiting for long procurement and infrastructure approval cycles. Relevance comes with high uncertainty at the beginning of both a startup and product-project development, when classic corporate data warehousing takes from six months up to two years to deploy, thus injecting great schedule and budget risk on top of this, reducing the speed of validated learning through loss of the value of data due to lack of operational feedback. It is new in the mixture of three-tier architecture (Staging, Transform, Data Mart), usage of any current DBMS or cloud engines (in temporary cluster mode, DuckDB, ClickHouse, BigQuery Sandbox), content analysis regarding availability of SQL skills and economic risk assessment together with a systematic comparative and instrumental analysis of performances of prototypes. The main finding is that the cycle from event arrival to target table update can fit within fifteen minutes which means fulfillment of a requirement that needs fast reaction for changes in user behavior and marketing campaigns while keeping the flexibility on the structure level among SQL prototypes preserving transparency and reproducibility plus automatic policies deleting obsolete data and serverless sandbox mode controlling costs. A smooth transition from the temporary solution to stable platforms, according to the Infrastructure as Code principle, minimizes operational risks and ensures continuity of metrics. The article will be helpful to startups and product teams, data engineers, and business analysts seeking to combine the speed of Lean methodology with data reliability
Third-Party Verification Badging for Consumer Trust: Case Study from Health and Nutrition Categories
Full text linkThe article examines the role of independent certification and trust-badge systems as an institutional mechanism for reducing information asymmetry in the health and nutrition segments. The relevance of the study is driven by the rapid growth of dietary supplements and functional products, accompanied by escalating consumer distrust and the rise of label skeptics. Under such conditions, external verification functions not as an optional marketing attribute but as a key quality signal that bridges costly auditing with the consumer’s simplified cognitive processing of information. The work aims to analyze the theoretical foundations and practical cases of independent badges in healthcare and nutraceuticals, as well as to assess their capacity to generate a durable trust premium. The novelty of the study lies in a comprehensive examination of certification through the lens of the economics of trust and cognitive psychology: the badge is interpreted simultaneously as a costly signal of a manufacturer’s probity and as a visual hash that facilitates consumer choice. The main results show that the display of an outward sign not only raises the subjective feeling of safety and lowers perceived risk but also translates into financial benefits, from increased readiness to pay to boosted loyalty and keeping a price premium. At the same time, the system’s weakness is revealed: the doubting of one mark can throw doubt on a whole setup, making protocol openness and reputation risk control key parts of certification strategy. The article will be helpful to researchers in the economics of trust, marketers, experts in health and nutrition, and practitioners engaged in designing and implementing independent certification systems
Enhancing Healthcare Data Interoperability with AI-Driven Synthetic Datasets Using FHIR Standards
Full text linkThis article addresses how one can combine AI-generated synthetic medical datasets as well as FHIR standard semantic standards and APIs to better enable cross-system interoperability between health care providers. In order to create data sets with the desired level of analytical utility, the author have as a study objective to demonstrate how the generation of synthetics in the context of FHIR resource and transport via the R5 mechanisms of profiles, subscriptions and asynchronous export to NDJSON removes both legal and technical barriers to accessing clinical data. Because data-sharing is complicated by the risk of regulatory noncompliance; a significant amount of clinical data is contained in unstructured documents; and clinical data grows at an exponential rate, the need for this type of solution has increased. This work is innovative in that it brings together three families of generative models: sequential LLM-like models, GANs and diffusion networks. Additionally, the author use a direct mapping pipeline from generative models to FHIR resources and validate profile IG and automated REST tests using built-in systems. To train the generative models on rare nosologies, and to perform load-testing, the author describe an architecture that enables private, secure access to multiple, representative cohorts, which are used to generate the synthetic data. Key Findings: (1) The use of synthetics inside the FHIR shell ensures compatibility with the current infrastructure while reducing legal barriers. (2) They also allow risk-free testing of extreme scenarios and (3) they also reduce sample bias. A rational and practical strategy for accelerating the implementation of analytics in clinics would include a strategy of generating synthetics in a continuous integration/continuous delivery (CI/CD) environment and (2) implementing mandatory validation through profile management. The author believe that the article will provide valuable assistance to medical AI developers, integration solution architects, IT service managers, and regulatory analysts
Application of Text Summarization on Text-Based Generative Adversarial Networks
Full text link In this project, we wish to convert long textual inputs into summarised text chunks and generate images describing the summarized text. This project aims to cultivate a model that can generate true-to-life images from summarized textual input using GAN. GANs aim to estimate and recreate the possible spread of real-world data samples and produce new pictures based on this distribution. This project offers an automated summarised text-to-image synthesis for creating images from written descriptions. The written descriptions serve as the GAN generator\u27s conditional intake. The first step in this synthesis is the use of Natural Language Processing to bring out keywords for summarizing. BART transformers are employed. This is then fed to the GAN network consisting of a generator and discriminator. This project used a pre-trained DALL-E mini model as the GAN architecture
Optimizing Energy Efficiency on Task Allocation for Cyber Foraging in a Transient Mobile Cloud System
Full text linkIn this research, we address the essential problem of achieving energy-efficient task allocation, which is a vital building block of cyber foraging on a transient mobile cloud. The goal is to minimize the total energy consumption for collaborative task executions among mobile devices in a multi-hop mesh network constructed on a mobile agent-based framework. Accordingly, we propose an energy-efficient task allocation problem formulation that takes into account the required restrictions. Next, we develop an optimal task allocation solution based on the modification of the Kuhn-Munkres algorithm by leveraging on the structural properties of the problem. We further evaluate the effectiveness of the suggested task allocation scheme through numerical study on a simulated system. The simulation reveals a performance gain on energy consumption reduction over other widely used task assignment algorithms