International Journal of Innovations in Science & Technology
Not a member yet
813 research outputs found
Sort by
Improving Software Security Through an LLM-Based Vulnerability Detection Model
Full text linkThe risks to modern digital infrastructures posed by software vulnerabilities are critical and include data breaches, unauthorized access, and losses in revenue. Although traditional static and dynamic analysis tools are effective in discovering vulnerability patterns, they are not able to recognize complex, context-dependent, logic-based, and security-embedded flaws that evolve within software systems. This research offers a Large Language Model-based Vulnerability Detection Model (LLM-VDM) focused on enhancing software security with intelligent, context-aware code analysis. Leveraging transformer-based architecture adapted to the Juliet, Big-Vul, and Devign benchmark datasets to assess the performance and integration of code semantic and code contextualization methods, the proposed model was evaluated. Experimental results demonstrated LLM-VDM’s superiority to both baseline and deep learning competitors SonarQube, Devign, CodeBERT, and CodeT5, attaining 91.2% accuracy, 90.0% F1-score, and 0.94 AUC. Furthermore, the integrated explainability module improves explainability by pinpointing vulnerable code and outlining remediation strategies. The findings showed LLM-based technology provides software developers with more secure, adaptive, explainable, and scalable systems, meeting the needs of contemporary software development
Microcontroller-Based Automated Cloth Folding Machine for Domestic and Industrial Textile Applications
Full text linkManual cloth folding is a repetitive and time-consuming process that limits productivity and efficiency in both domestic and industrial contexts. This work presents the design and implementation of a microcontroller-based automatic T-shirt folding machine aimed at achieving a fully autonomous operation without manual intervention. The proposed system employs lightweight acrylic boards for structural strength and portability. It integrates sensors to ensure precise garment detection and controlled folding operations. Experimental evaluation demonstrates a folding time of 9.2 seconds per shirt, ensuring operational efficiency and precision. The system effectively addresses the limitations of previous designs, particularly in terms of automation, sensing capability, and user safety, and provides a scalable foundation for future advancements in automated garment handling
Education Access System (EAS): A Low-Latency Learning System for Low-Bandwidth Education Access
Full text linkOnline learning worldwide requires well-performing Learning Management Systems (LMS); however, most web-based application packages have no opportunities to execute their task in the low-bandwidth areas due to the infrastructure factor, the absence of digital literacy, and the barriers of institutions. It applies this digital divide to the countryside and the developing world and limits access to quality education. Existing LMS systems such as Moodle and Blackboard are designed to support high-bandwidth systems, and they are not applicable in locations where 2G/3G is the predominant network. This gap can be addressed by a low-bandwidth, lightweight, and low-latency LMS named Education Access System (EAS), which is developed, engineered, and tested in this research project. It also worked through four phases of methodology: requirement analysis and stakeholder mapping, data analysis and system architecture built upon a layered architecture, implementation of the application using React, node.js, MongoDB, and optimized APIs, and assessment of the project via manual/automated testing and User Acceptance Testing (UAT) with students. The findings show login and usable databases, 95.8% respondents had improved learning with the platform, but valid performance under the simulated conditions of a 2G/3G (100-500 kbps) was still simulated (low forum interaction 12.5%). EAS explains how LMS optimization solutions can be used to the advantage of equitable digital education, comprising SDG 4, and how the future will center on mobile deployment, offline adaptations, and long-term collaborative capabilities
A Data-Driven Review of Machine Learning Techniques for E-commerce Product Recommendation Systems
Full text linkIn today’s digital economy, recommendation systems are essential for enhancing customer experience and driving e-commerce growth. This study presents a comparative, quality-ranked review of machine learning-based product recommendation techniques, evaluating five key approaches: association rule mining, content-based filtering, collaborative filtering, knowledge-based systems, and hybrid models. Using a systematic literature review of 44 peer-reviewed publications across major publishers, the analysis includes geographic and publisher-wise trends and a structured quality assessment rubric. Results highlight hybrid systems as the most promising strategy, offering superior accuracy, diversity, and personalization while addressing cold-start, sparsity, and scalability challenges. Each technique’s strengths, limitations, and practical deployment considerations are critically examined to support evidence-based decision-making. The study concludes by recommending hybrid approaches tailored to domain-specific needs, offering actionable insights for both researchers and industry practitioners seeking effective and adaptable recommendation systems
Tracking Temporal Migration of the Indus River: Morphological Changes in a Downstream Reach
Full text linkIndus River morphological changes create environmental challenges, impacting local communities and ecosystems through fertile land loss, bank erosion, and higher flood risks. Monitoring these changes is crucial for flood and water resource management and infrastructure protection. This study uses geospatial data and tools to analyze spatial and temporal morphological dynamics of a downstream Indus River reach between the Sukkur and Kotri barrages from 1995 to 2024. Satellite imagery was analyzed to study morphological changes. Significant channel adjustments in river shape and form were observed, evident through erosion, deposition, and lateral shifts over the past three decades. The maximum erosion, covering 35,540 ha, and accretion, covering 23,737 ha, were observed between 1995 and 2005, with later periods showing reduced erosion and greater stability. The total cumulative erosion was 71,575 ha, and the total cumulative accretion was 64,790 ha, which gives a net loss of 6,785 ha. The sinuosity analysis showed that the meandering tendency of the river increased over the years as the sinuosity ratio increased from 1.82 in 1995 to 1.93 in 2024. These findings reveal the features of fluvial dynamics of the Indus River and stress the importance of reducing the adverse effects of these changes as necessary for the area\u27s sustainable development
Assessing the Efficacy of Pixel-based and Object-based Classification Techniques and Classifiers for Land Cover Mapping Using Landsat-8 and Sentinel-2 Data in Complex Mountainous Terrain
Full text linkDisaster mitigation and climate-resilient planning heavily depend on accurate Land Use and Land Cover (LULC) datasets. Well-classified LULC data optimizes hazard modeling, surface runoff estimation, and sustainable land use planning, enabling informed decision-making and proactive risk reduction. However, supervised LULC classification faces challenges such as selecting optimal Machine Learning (ML) algorithms, differences in spatial and spectral resolution, and seasonal variability. This study adopts a multi-tiered approach to generate effective LULC maps for Gilgit District, Pakistan, by comparing pixel-based classification and object-based image analysis (OBIA) methods. Pixel-based classification was performed on Google Earth Engine (GEE) using Landsat-8 and Sentinel-2 imagery, applying three classifiers: Random Forest (RF), Support Vector Machine (SVM), and k-Nearest Neighbor (k-NN). OBIA involved multi-resolution segmentation, followed by training and classification on image objects using the same algorithms. Validation using independent samples revealed that object-based maps were visually smoother and more realistic. Quantitatively, pixel-based RF yielded the highest accuracy: 82.9% for Landsat-8 and 78.02% for Sentinel-2. In contrast, OBIA k-NN achieved superior accuracy: 81.3% on Landsat-8 and 83.6% on Sentinel-2. Remaining classifiers also provided nearby results in both classification methods. Lower accuracy in Sentinel-2 may be due to within-class spectral variability at 10m spatial resolution, while Landsat-8’s lower resolution (30m) reduced object-based segmentation performance, resulting in object heterogeneity and misclassification. Although pixel-based classification provided promising results, OBIA ultimately demonstrated superior overall accuracy. This study highlights the importance of resolution-context compatibility and algorithm choice in enhancing LULC classification, which is essential for reliable climate-responsive planning, disaster preparedness, and sustainable development
Floods and Flood Hazard Assessment in the Floodplain of River Swat, District Charsadda, Pakistan
Full text linkThis study aims to evaluate flood risks and carry out flood hazard assessment in the floodplain of the River Swat, District Charsadda. This study focuses on two objectives: primarily, to explore the flood situation in the study area, and secondly, to carry out the flood hazard assessment in the floodplain of the River Swat, District Charsadda. The study area is located 27 km north-east of Peshawar city. District Charsadda is part of the Peshawar valley, and the study area covers an area of 1,593 Km2. The gentle slope is from north to south, which plays a major role in making it vulnerable to recurrent flood events. In District Charsadda, the floodplain of the River Swat is highly vulnerable to recurrent flooding during the summer season. There is a lack of flood hazard assessment for management strategies. Therefore, to minimize the negative impacts of floods, flood hazard assessment and management strategies will help reduce the losses resulting from recurrent flooding. There is a need to identify the flood hazard trend in the study area and to generate the flood hazard zonation map. The data are collected from PMD, PIDA, WAPDA, and the Survey of Pakistan. A SPOT recent 2.5m resolution satellite image is used for land use data, and the SRTM data of 90m resolution is used for the generation of a contour map, DEM, and drainage pattern. From the collected data, a hydrograph is created and projected, and the frequency of flood recurrence is determined. To obtain a good picture of the occurrence of floods, the data were also connected with temperature and rainfall characteristics. The AHP method is used to develop a flood hazard zonation map, for which the probability and recurrent intervals of the flood hazard are calculated by using the 24-year data from 1998-2022, and graphed. This clearly shows the recurrence of the flood hazard with specific magnitudes. The physical parameters, including the discharge, amount of rainfall, and elevation data, are used to develop a flood hazard zonation map under a combination of five zones
Vermicompost Effects on Carbon Sequestration of Paulownia Elongata in Agroforestry System
Full text linkIn a controlled agroforestry system, this study evaluated the effects of vermicompost-silt mixtures (30:70, 40:60, and 50:50 ratios) on the growth and carbon sequestration potential of hybrid Paulownia elongata. Compared to other ratios and the control, the 50:50 treatment significantly (p ≤ 0.05) increased plant height (25 inches), stem diameter (0.87 cm), and biomass (13.22 g/plant). According to soil analysis, the 50:50 mixture had the highest carbon stock (6.61 g/plant) and improved potassium (582.67 mg/kg) and organic matter (1.12%) contents. These findings show that Paulownia growth and carbon capture are maximized by balanced vermicompost application, providing a sustainable method for agroforestry in nutrient-deficient soils
Voice Cloning and Synthesis Using Deep Learning: A Comprehensive Study
Full text linkThis paper reviews current voice cloning and speech synthesis methods. It focuses on the way that deep learning enhances AI-generated voice synthesis in terms of quality, flexibility, and efficiency. We analyze the top AI models in terms of their significance to virtual assistants, dubbing, and accessibility tools: XTTS_v2, Whisper, and Llama 8B. Voice cloning and TTS efforts in Tortoise are improved by XTTs_v2. Based on the multilingual creative transfer, it has a higher speed and shorter time of a computational process, and generates synthetic speech closer to naturalness. Whisper is a transcription model that goes from an audio waveform to text. It simplifies access to audio data. Llama 8B focuses on user question answering for enhancing AI and human interaction. Other related work includes fastSpeech2 [1], Neural Voice Cloning with few Samples [2], and Deep Learning-Based Expressive Speech Synthesis [3], which also contribute to these advancements. This progress enhances machines\u27 ability to communicate in an emotional and human-like way, leading to more sophisticated technology
MatLab Bvp4c Technique to Compute Thermophoresis and Brownian Motion in Nanofluid Flow Over a Transient Stretching Sheet
Full text linkThis physical phenomenon examined the transport mechanisms of heat and mass within a nanofluid thin film. The nanofluid thin film is situated over an unsteady stretching sheet, which is one of the pioneering contributions to the field, focusing specifically on the flow dynamics of nanofluid thin films. This foundational framework is established by Buongiorno’s fluid model. The mathematical model is applied for the evaluation of the nanofluid film, which adeptly weaves in significant phenomena, including Brownian motion as well as thermophoresis. The mathematical model is achieved in the form of non-linear partial differential equations (PDEs) for computation with the help of computer applications. Firstly, the analytical framework of similarity transformations is applied to non-linear PDEs to convert them into ordinary differential equations (ODEs). Secondly, these ODEs have been critically examined and prepared for coding in MatLab by reducing their high order into first order. The software Mathematica and MatLab have been employed to solve the boundary value problem (BVP). The built-in BVP4c solver is applied to obtain accurate solutions in the form of graphs and numerical values. The current analysis yields significant results revealing that both the free surface temperature and the volume fraction of nanoparticles tend to increase in response to variations in both unsteady conditions and magnetic parameters. Furthermore, the outcomes demonstrate that the interaction among diverse nanofluid variables with the phenomenon of viscous energy loss contributes to a reduction in the overall heat transfer rate. The potential effect of these proficient thermal management techniques is crucial, especially in microelectronics and energy systems