International Research Journal of Multidisciplinary Technovation

Bi2O3-Incorporated V2O5 Nanocomposite: Enhanced Structural, Optical, and Electrochemical Properties for Supercapacitor Applications

Jency Sebatine P — 2026-06-08

The present work is an attempt to explore the impact of bismuth oxide (Bi₂O₃) at different concentrations (5%, 10%, and 15%) on the structural, optical, and electrochemical features of hydrothermally synthesized V₂O₅ nanoparticles. Structure characterization via X-ray diffraction (XRD) confirmed that the orthorhombic phase of V₂O₅ remains in all composite samples, with slight changes caused by Bi₂O₃. The study of optical properties involved DRS-UV, and the results pointed to the band-gap gradual increase, notably for 15 % Bi₂O₃sample. The shape changes as well as the particle distribution changes due to the composite formation were confirmed by FE-SEM-EDX. XPS study showed that the pure V₂O₅ chemical environment had been changed after the addition of Bi2O3. Electrochemical evaluations, achieved by cyclic voltammetry and charge-discharge cycling, revealed considerable improvements in capacitance as well as stability. The 15% Bi₂O₃/V₂O₅ nanocomposite electrode demonstrated exceptional cycling durability with the largest specific capacitance (487 F g⁻¹ at 1 A g⁻¹), and outstanding retention of 89% after 5000 cycles among the investigated concentrations. The assembled asymmetric device with 15% Bi₂O₃/V₂O₅ nanocomposite and activated carbon attained an exceptional energy density of 38.4 Wh kg⁻¹ and a power density of 16,000 W kg⁻¹.

A Modified Genetic Algorithm–Based Feature Optimization Framework for Cardiovascular Disease Risk Prediction

Shamal Salunkhe — 2026-07-07

The proper diagnosis of cardiovascular disease (CVD) risk rises as a significant issue in clinical practice because the risk is directly associated with patient survival. Thus, correct and timely assessment of the risk is crucial. This work suggests a CVD probability prediction model, which combines ensemble learning and Deep Learning (DL) algorithms with a feature choice approach supported by a Modified Genetic Algorithm (MGA). A balanced clinical dataset consisting of 1,025 patient records and 14 medically relevant attributes was used which was retrieved via a publicly available repository with benchmarks dataset. The comparison of a broad range of predictive models, including classical Machine Learning (ML) algorithms like Logistic Regression, Support Vector Machines (SVM), K-Nearest Neighbors, Naïve Bayes, Decision Trees, Random Forests, XGBoost, and LightGBM, as well as DL models, including Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), and Multi-layer Perceptrons (MLP), was made. The MGA based feature collection is used to select the most informative clinical attributes, decrease feature redundancy, and increase discriminative ability. The outcomes of experimental findings prove that the models with the assistance of MGA are most effective as they are more predictive and stable across various classifiers. The results demonstrate the imperative need of intelligent feature selection to enhance model generalization and predictability, which reinforces predictive modelling framework with feature optimization against cardiovascular risk prediction. The paper presents the framework that mediates between optimization methods and predictive modeling to provide valuable information on the next generation of data-driven cardiovascular diagnostics.

Blockchain-Enabled E-Governance Framework for Secure and Transparent Land Record Management

Aseem Khanna — 2026-07-04

Land disputes form a substantial share of pending civil cases in India, and the cost is structural: ownership records sit across five separate offices patwari, tehsildar, sub-registrar, survey, and State Revenue Department that have never shared a single source of truth. Existing blockchain proposals for Indian land records either treat the chain as a future replacement for the statutory record, which is legally premature, or build flat permissioned ledgers that ignore the role hierarchy governing the actual workflow. This paper presents a prototype that addresses this gap by modelling role-aware governance, privacy-preserving document handling, and auditable workflow execution within a Hyperledger Fabric-based land-record framework. The proposed system models six administrative roles as distinct Membership Service Providers on Hyperledger Fabric v2.5.4, encodes mutations, encumbrances, and disputes as first-class chaincode states with role-aware endorsement, and combines ciphertext-policy attribute-based encryption for long-lived role-driven document access with identity-based encryption for short identity-and-time-bound disclosures. This 2 in 1 privacy design increases the level of confidentiality and enables administrators to easily track users' activity at any time as needed. With the workload mix applied using publicly released sub-registrar volume data of two districts of Punjab, the benchmark attained a peak throughput of 612 ± 18 tps and had a mean end-to-end latency of 1.84 s at 400 tps on a cluster of 5 nodes. The observed saturation knee was blamed on CouchDB write contention, as opposed to ordering or endorsement being a problem. The framework is important, as it is proposed as a validation and audit overlay to be used in parallel with the statutory record.

Attention Based Swarm-optimized Quantum CNN for Sentiment Analysis in E-commerce Products

Gokilavani A — 2026-07-08

In this study, a hybrid intelligent model based on deep learning techniques is presented, which exploits feature representation, optimization and enhanced feature learning capabilities. Quantum inspired Convolutional Neural Network, Efficient Dynamic Transformer, and Efficient Low-Rank Attention methods are incorporated into the design of the model to boost feature learning capabilities, context modeling capabilities and computational efficiency. First, the input data are cleaned and normalized to eliminate noise and normalize feature values. Quantum CNN is then utilized to generate feature representations, providing more efficient discriminability when compared with the traditional convolution approach. The extracted features are then used by the Efficient Dynamic Transformer to capture long-term contextual interactions, whereas Efficient Low-Rank Attention minimizes computation overheads through low-rank approximations. A series of experiments was carried out on benchmark datasets to evaluate the effective performance of the proposed method. The developed model yielded a mean classification accuracy rate of 99.47%, demonstrating superior performance to multiple state-of-the-art baseline models, including CNN, LSTM, and Transformer-based models. Other measures, such as precision, recall, F1 score and kappa score further validate the effectiveness of the proposed framework. Furthermore, experimental results show that the developed sentiment analysis model exhibits enhanced computational efficiency and generalization capabilities compared to current algorithms.

Impact of Traffic Congestion on Fuel Consumption and Emissions: A Baghdad Case Study

Esraa R. Al-gurah — 2026-06-21

Traffic congestion in fast-growing cities results in considerable growth in fuel usage and emission by vehicles, especially under slow driving conditions or where there are frequent stops and starts in the movement. This research uses simulation-based case studies to examine the effect of traffic congestion in Baghdad city with regard to the impact on fuel consumption and vehicle emissions. It is important to note that an analytical model has been developed in this study, where the connection between traffic, fuel usage, and emissions has been modeled with the help of inventory-based fuel emissions model. Three main levels of traffic congestion were considered, as well as some other traffic states, to study the differences in average speed, fuel consumption, and emissions. According to the study findings, the higher level of traffic congestion the more fuel is consumed, which directly translates to a proportionate increase in emissions as a result of the chosen methodology of estimating the level of emissions.

A Novel Deep Learning Framework for Automated Brain Tumor Classification Based on CNN-BiLSTM Model

Chembeti Saraswathi — 2026-07-07

Early identification of brain tumors is crucial for effective therapy and improved clinical outcomes. For brain tumor detection, this study presents a novel neural network that combines a three-layer Convolutional Neural Network (CNN) with a Bidirectional LSTM (BiLSTM). The proposed approach can automatically identify complex patterns and features from medical imaging data by using advanced computational techniques, leading to faster and more accurate evaluations. This study contributes to the field of medical image analysis by providing a reliable and accurate computational approach for brain tumor recognition. The BR35H dataset is used in this research, which comprises MRI images of both tumor and non-tumor subjects. The proposed CNN-BiLSTM model showed strong performance in tumor detection, with an accuracy of 98%, a precision of 98.48%, a recall of 97.5%, and an F1-score of 97.99%. The effectiveness of the proposed method is demonstrated through a thorough final evaluation, which includes a confusion matrix, comparison with current algorithms, and experiments using different data sizes. These experimental outcomes show the significance of combining CNN and BiLSTM models for brain tumor recognition.

Artificial Rabbits Optimized Fuzzy Elliptic Curve Signcryption for Secured Data Transmission in WSN

Paruvathavardhini J — 2026-06-08

Secure and energy-efficient data transmission is a significant challenge in Wireless Sensor Networks (WSN). Many energy-saving and security schemes already address this issue. However, these schemes fail to strike a balance between energy efficiency, performance, and security, making them unsuitable for WSNs with limited resources. This research proposes a new system for energy-efficient and secure data transfer in WSNs to enhance network reliability and lifetime. The scheme has two phases: forwarder node selection and secure data transfer. In the selection phase, a Gaussian Likelihood Censored Regression (GLCR) method selects the most efficient adjacent node as the forwarder node, based on factors such as energy, packet loss, and others. This continues until the sink node is reached, creating a reliable path. Once a path is set, secure data transfer occurs using a proposed Optimized Fuzzy Elliptic Curve-based Signcryption (OFECS) scheme. Here, the Artificial Rabbits Optimization Algorithm (AROA) fine-tunes Fuzzy Elliptic Curve (FEC) parameters like elliptic curve coefficients, cofactor, group order, generator point, and prime number. This process enhances cryptographic security while reducing computational overhead. The scheme is simulated using the NS2 tool. Finally, results show that GLCR-OFECS achieves a 99.3% Packet Delivery Ratio (PDR), 0.8 mJ of Energy Consumption (EC), 1.3 ms of End-to-End Delay (EED), 245 kbps of throughput, 512 seconds of network lifetime outperforming existing schemes. For 2000 data units, it achieves 98.57% data confidentiality, 97.74% data integrity, 26.1 MB space complexity, and 32 ms execution time, surpassing existing WSN schemes for data transfer.

IPL: An Intelligent, Predefined, and Lightweight Recovery Scheme for Node Failures in Topology-Aware Software-Defined Wireless Sensor Networks

Sathish S — 2026-07-04

Software-defined wireless sensor networks (SDWSNs) improve network programmability and centralized control, but maintaining connectivity under node and link failures remains difficult because of node mobility, limited energy, and dynamic topology changes. This study proposes IPL, an Intelligent, Predefined, and Lightweight recovery framework for topology-aware SDWSNs. The framework integrates three coordinated mechanisms: predictive link-lifetime estimation using energy and mobility parameters, energy-aware target positioning through a weighted midpoint strategy, and ring-based coordination among mobile IPL relay nodes for deterministic and low-overhead recovery. IPL is designed to handle both isolated and multiple concurrent failures while reducing controller burden and avoiding expensive global recomputation. The method was evaluated in a Mininet/Floodlight-based SDWSN environment with 150 nodes under identical settings against four benchmark schemes: IFT, Fed-TSN, P4Neighbor, and LCD. Across varying failure conditions, IPL consistently achieved faster recovery and better communication reliability. Relative to the baselines, the proposed method reduced recovery time by up to 26%, lowered latency by up to 27%, decreased energy consumption by up to 18%, improved packet delivery ratio by up to 19%, increased recovery success rate by up to 17%, and extended network lifetime by up to 19%. These gains arise from proactive link monitoring, rapid relay repositioning, and structured recovery coordination. Overall, IPL offers an efficient and scalable recovery solution for dynamic SDWSNs, particularly in environments with moderate failure rates, while highlighting opportunities for future enhancement through adaptive relay allocation and improved mobility-aware prediction.

Improved Hybrid Model-Based Machine and Deep Learning Approach for Intrusion Detection System

Premananda Sahu — 2026-07-08

The Intrusion Detection Systems (IDSs) are very important tools for defending a network against emerging cyber threats. This study proposes the hybrid intrusion detection system model of extreme gradient boosting (XGBoost) and KMeans++ clustering algorithm to balance the trade-off between accuracy, efficiency, and robustness in detecting malicious traffic. XGBoost algorithms are good for structured problems where classification problems occur, whereas KMeans++ helps you to get more clustering accuracy by helping centroid initialization. For enhancing the performance of the model some feature extraction steps and data preprocessing steps like normalization, encoding, Synthetic Minority Over-sampling Technique (SMOTE) based imbalance data balancing were considered. The system was trained and validated on Canadian Institute for Cybersecurity Intrusion Detection System (CICIDS) 2017 and put to test in various metrics: accuracy, precision, recall, F1-score, ROC-AUC and false positive rate. Comparative analysis was performed using traditional machine learning models such as SVM, Decision Tree, Random Forest, Naive Bayes and deep learning architectures which include CNN, LSTM and Auto-Encoder. It was found to be high detection accuracy of 99.87% with very low FPR by far i.e. 0.1%. The model provided high recall and precision in different types of attack and successful overfitting resistance could be confirmed using 10-fold cross-validation, XGBoost regularization and structure clustering. This work shall play an important role in improving hybrid models to minimize alert fatigue with trustworthy threat classification in the real operational traffic.

Phytochemical Profiling, Antioxidant Potential, and Cytotoxic Activity of Hydroethanolic Extract from Prosopis cineraria Seeds

Sivasankar Murugesh — 2026-06-29

Prosopis cineraria (L.) is a perennial desert tree belonging to the family Fabaceae. It is widespread in India, the United Arab Emirates, Pakistan and Iran. The study aimed to evaluate the phytochemical screening, antioxidant potential, and to assess cytotoxic activity through a Brine Shrimp Lethality Assay. Artemia salina was used as a biological model to detect the toxicity of bioactive constituents. The qualitative analysis of the hydroethanolic extract of Prosopis cineraria seeds confirmed the presence of major phytoconstituents such as alkaloids, flavonoids, phenols, cardiac glycosides, anthraquinones, coumarins, and glycosides. FTIR spectroscopy showed dominant functional groups, including hydroxyl, amine, carbonyl, alkene and aromatic. Similarly, GC-MS chromatogram reveals the abundant compounds, such as 9, 12-octadecadienoic acid (linoleic acid), n-hexadecanoic acid, 3-O-methyl-D-glucose, Methyl salicylate, campesterol, and stigmasterol. The antioxidant potential of the extract was evidenced by IC₅₀ values of 46.77 µg/mL in the ABTS radical scavenging assay and 54.61 µg/mL in the hydroxyl radical scavenging assay. This indicates a significant capacity to neutralize free radicals, and the observed strong activity exhibited a concentration-dependent pattern. The brine shrimp lethality assay demonstrated significant concentration-dependent toxicity, with an LC₅₀ value of 930.3 µg/mL, indicating biologically active low-to-moderate cytotoxic potential (R² = 0.9972; F (4, 10) = 506.64, p < 0.0001).

Feature Importance of ERP Implementation Impediments: An Analysis Using Random Forest Model

Abhinandan M.S — 2026-07-07

The present research aims to assess the feature importance of Enterprise Resource Planning (ERP) impediments for implementation and to classify usage intention. To achieve the objective, responses were collected from 750 users. Classification model was designed using an RF algorithm which was developed for classifying intention to use the ERP system. The model has achieved 83% precision in predicting intention to use. Recall value and F1-score were noted as 0.67 and 0.74, respectively. The feature importance of Implementation Impediments of ERP systems was elicited through a correlation analysis between various features among the Implementation Impediments of ERP systems. Further, ROC curve establishes an AUC value of 0.92, representing highest discriminative power. The CV accuracy is noted as 0.798. The study outcome Firstly, it helps to explore feature importance impediments in the implementation process of ERP to comprehend and tailor the ERP features system itself drives the intention to use to enhance business performance, in turn, user satisfaction. Secondly, comprehending critical impediments and usage intentions of the system helps the organisations to effectiveness of system usage, which impacts user satisfaction. Thirdly, by integrating AI/ML, this research contributes by deploying an interpretable framework that offers data-driven decisions for planning, implementation, and evaluation of ERP performance. Fourthly, the outcome of this study is valuable to design and develop implementation strategies suitable for ERP users, which enhance stakeholders’ engagement and foster sustainable growth. By integrating AI/ML, this research contributes by deploying an interpretable framework that offers data-driven decisions for design, application, and evaluation of system performance. Helps to comprehend critical impediments and to enhance usage intentions of the ERP system. Also, offers data-driven decisions for ERP successful implementation.