25 November 2024 in Internet Technology Letters
6 December 2025 in International Journal of Environment, Engineering and Education
In today’s digital environment, online reviews have become one of the key factors that influence the decisions of customers. This is especially true in areas such as e-commerce, travel and the hospitality industry, where buyers depend heavily on the shared experiences of others before making a choice. At the same time, the growing issue of fake or fabricated reviews has raised serious concerns, as it reduces the reliability of online platforms and creates confusion for consumers. Detecting such misleading reviews is not an easy task, since the language used in them is often very close to what is seen in genuine opinions. In the present work, an attempt has been made to compare the performance of traditional machine learning techniques with that of transformer-based deep learning models for the identification of fake reviews. As part of the baseline, Logistic Regression and Linear SVM were applied with TF-IDF features. On the other hand, advanced architectures like BERT, RoBERTa and XLNet were fine-tuned on the Deceptive Opinion Spam Corpus. The results clearly indicated that the classical models gave accuracies in the range of mid-80 percent, whereas the transformer-based models performed much better, crossing or coming close to 90 percent. Among the transformer models, RoBERTa showed the most balanced performance across precision and recall, XLNet gave the highest recall, which is very important when sensitivity is the main concern, while BERT achieved competitive results with less demand on computing resources.
1 June 2018 in Journal of Computational and Theoretical Nanoscience
In the Linked Open Data, there is a tendency that connected datasets will change often. These changes directly lead to updating on a distinctive attribute and address of entities. The address modification in Resource Description Framework elements makes their relating joins to be broken. A Broken link is one of the real impediments that the LOD is confronting. Most methods to deal with this issue endeavour to settle broken connections at the link destination point. These methodologies have two noteworthy issues: will stop the entire system from working (SPOF); and dependence on the link destination data source. In this paper, we present a strategy for settling broken links which depends on the link source point, and find the new address of the disconnected entity. To the source point, we present two datasets, which we call “Point-To” and “Point-From.” Using these datasets, our strategy forms a cliquish graph structure for every entity that should be monitored often. This structure is utilized to recognize and find the new address of the isolated entity. Thereafter, the most comparative entity, which is a contender for the isolated entity, is concluded and proposed by the algorithm. The proposed paragon is assessed with DBpedia dataset under the domain of “person” entities. The outcome demonstrates that the greater part of the broken links, which had alluded to a “person” entity in DBpedia, had been fixed accurately.
1 August 2026 in IAES International Journal of Artificial Intelligence (IJ-AI)
Distributed denial of service (DDoS) attacks continues to pose serious risks to modern networks, with their growing intensity making early detection both critical and challenging. Conventional machine learning (ML) models often struggle with the nonlinear and highly dynamic nature of attack traffic, which motivates the use of advanced architectures. In this study investigate a transformer-based classifier for DDoS detection on the CIC-DDoS2019 dataset. The workflow included preprocessing, feature scaling, and domain-guided feature selection. Logistic regression (LR) was employed as a baseline, achieving 92.1% accuracy and F1-score of 0.90, thereby revealing the limitations of linear models. The transformer, after hyperparameter tuning and 5-fold cross-validation, reached an average accuracy of 99.95% with precision, recall, and F1-scores all above 99.9%. The model demonstrated stable convergence and generalization across folds. These results highlight the strength of attention mechanisms in capturing feature dependencies, while also pointing to future directions such as real-time deployment, explainability, and resilience to zero-day attacks.
24 September 2026 in International Journal of Computational Intelligence Systems
Energy efficiency that potentially handles the techniques of routing and clustering is considered as the most significant design concern in Wireless Sensor Networks (WSNs). This energy efficiency in WSNs is required for attaining stable routing process and extending network lifetime. In this paper, Multi-Objective Hybrid Chimp and Naked Mole Rat Optimization Algorithm (MHCNMROA)-based cluster routing protocol is propounded for concentrating on process of facilitating energy efficient cluster routing that selects optimal Cluster Head (CH) selection for prolonging network lifespan. This energy efficient protocol initially used Chimp Optimization Algorithm (COA) for ideal CH selection by considering factors like node centrality, distance to neighbours, Residual Energy (RE), distance to Base Station (BS) and node degree during fitness evaluation of sensor nodes. It further used fuzzy NMROA for achieving routing between BS and chosen CHs in the network. This utilization of fuzzy decision module helps in dynamic adjustment in the searching behaviour of NMROA. It adopted fuzzy rules that are related to the parameters associated with distance to BS, RE and node degree for handling constraints of NMROA during routing. It finally attained cluster maintenance for preventing the failure of nodes such that network lifetime is improved during data transmission by establishing best path between source and BS. The experimental results of MHCNMROA confirm its efficacy in maximizing Packet Delivery Ratio (PDR) and throughput with reduced energy consumption and packet drop on par with the competitive protocols.