Bennett University

The information deployment on social networks through word-of-mouth spreading by online users has contributed well to forming opinions, social groups, and connections. This process of information deployment is known as information diffusion. Its process and models play a significant role in social network analysis. Seeing this importance, the present paper focuses on the process, model, deployment, and applications of information diffusion analysis. First, this article discusses the background of the diffusion process, such as process, components, and models. Next, information deployment in social networks and their application have been discussed. A comparative analysis of literature corresponding to applications like influence maximization, link prediction, and community detection is presented. A brief description of performative evaluation metrics is illustrated. Current research challenges and the future direction of information diffusion analysis regarding social network applications have been discussed. In addition, some open problems of information diffusion for social network analysis are also presented.

One of the most fatal diseases that affect people is skin cancer. Because nevus and melanoma lesions are so similar and there is a high likelihood of false negative diagnoses challenges in hospitals. The aim of this paper is to propose and develop a technique to classify type of skin cancer with high accuracy using minimal resources and lightweight federated transfer learning models. Here minimal resource based pre-trained deep learning models including EfficientNetV2S, EfficientNetB3, ResNet50, and NasNetMobile have been used to apply transfer learning on data of shape$\:\:224\times\:224\times\:3$. To compare with applied minimal resource transfer learning, same methodology has been applied using best identified model i.e. EfficientNetV2S for images of shape$\:\:32\times\:32\times\:3$. The identified minimal and lightweight resource based EfficientNetV2S with images of shape $\:32\times\:32\times\:3$ have been applied for federated learning ecosystem. Both, identically and non-identically distributed datasets of shape $\:32\times\:32\times\:3$ have been applied and analyzed through federated learning implementations. The results have been analyzed to show the impact of low-pixel images with non-identical distributions over clients using parameters such as accuracy, precision, recall and categorical losses. The classification of skin cancer shows an accuracy of IID 89.83% and Non-IID 90.64%.

This study explores the synthesis and dynamic performance of polylactic acid (PLA) composites reinforced with bio-fibers, specifically jute and flax, combined with particulate fillers such as coconut shell and eggshell powder. The objective is to evaluate the viability of these composites as eco-friendly alternatives to conventional synthetic materials in structural applications. Through a series of mechanical tests, including tensile, flexural, impact, hardness, creep, and fatigue assessments, the enhanced mechanical properties and long-term durability of the reinforced PLA composites are demonstrated. Results indicate that jute-reinforced composites, particularly those containing eggshell or coconut shell powder, exhibit superior tensile strength, hardness, and impact resistance, making them suitable for applications requiring both strength and toughness. Flax-reinforced composites provide a balance of strength and flexibility, while the inclusion of particulate fillers improves stiffness and durability. Creep and fatigue testing further reveal that jute and flax composites exhibit enhanced resistance to long-term deformation and cyclic loading, with the JPC (jute/PLA/coconut shell powder) composite showing exceptional performance. Among the composites, the JPC composite (jute/PLA/coconut shell powder) had the maximum surface roughness of 0.074, while the FPE composite (flax/PLA/eggshell powder) had the least roughness value of 0.0038. The JPC composite (jute/PLA/coconut shell powder) demonstrated the highest hardness score of 99 Shore D, and the JPE composite (jute/PLA/eggshell powder) also scored relatively high at 97 Shore D. These findings suggest that natural fiber-reinforced PLA composites offer a sustainable, cost-effective solution for industries seeking materials with both environmental and mechanical benefits. Future research should focus on optimizing fiber-matrix interactions to enhance the performance of these composites in demanding applications.

Our present and evolved understanding has challenged the previously synonymous use of the terms ‘sex’ and ‘gender’. We have moved beyond the binary categorization towards proliferation of gender identities. Thus, raising questions whether certain identities are traits or gender identities. This complexity of the issue is exacerbated by the cultural relativity of gender identity and by lack of a standardized list. Adopting a balanced approach, the article touches upon the prejudices against the gender minorities. Additionally, it touches upon the controversies surrounding gender identities and their development to ensure that the individuals can make more informed choices and engage in more meaningful discourse. It addresses issues of politicised debates, linguistic diversity, and the role of pharma industry in the sex-change procedures. The paradigm of gender has transcended the binary constructs in the contemporary discourse. However, it has ventured into unchartered territories revealing several unexplored facets that await scholarly investigation. The present paper critiques the concept of gender identities and the sociopolitical landscape surrounding it through the lens of Critical Theory. In conclusion, our understanding of gender is still limited and evolving. There is a need for adopting a more nuanced and informed approach to challenge the issues posed by this era of evolving gender expression and identities. The article concludes with policy recommendations based on insights gained from the article and suggestions for future research.

In the current cybersecurity landscape, Distributed Denial of Service (DDoS) attacks have become a prevalent form of cybercrime. These attacks are relatively easy to execute but can cause significant disruption and damage to targeted systems and networks. Generally, attackers perform it to make reprisal but sometimes this issue can be authentic also. In this paper basically conversed about some deep learning models that will hand over a descent accuracy in prediction of DDoS attacks. This study evaluates various models, including Vanilla LSTM, Stacked LSTM, Deep Neural Networks (DNN), and other machine learning models such as Random Forest, AdaBoost, and Gaussian Naive Bayes to determine the DDoS attack along with comparing these approaches as well as perceiving which one is about to give elegant outcomes in prediction. The rationale for selecting Long Short-Term Memory (LSTM) networks for evaluation in our study is based on their proven effectiveness in modeling sequential and time-series data, which are inherent characteristics of network traffic and cybersecurity data. Here, a benchmark dataset named CICDDoS2019 is used that contains 88 features from which a handful (22) convenient features are extracted further deep learning models are applied. The result that is acquired here is significantly better than available techniques those are attainable in this context by using Machine Learning models, data mining techniques and some IOT based approaches. It’s not possible to completely avoid your server from these threats but by applying discussed techniques in the present juncture, these attacks can be prevented to an extent and it will also help to server to fulfil the genuine requests instead of sticking in the accomplishing the requests created by the unauthentic user.

This study aims to investigate the current landscape of immersive technologies and metaverse literacy among librarians in the Gulf region, exploring their familiarity, perceptions, and potential impact on library services. A cross-sectional design was adopted, utilizing a comprehensive survey questionnaire distributed to library professionals in six Gulf countries. The survey collected data on participants' familiarity with immersive technologies, perceived importance of metaverse literacy, and barriers to adoption. The findings indicate a significant level of familiarity among librarians with immersive technologies such as virtual reality (VR), augmented reality (AR), and virtual environments (VEs). There is a consensus among participants regarding the importance of metaverse literacy for providing relevant and up-to-date library services. Despite this recognition, various barriers to adoption were identified, including cost, lack of expertise, privacy concerns, and technical limitations.

In the intricate complexities of modern financial markets, the capability to predict stock values with high accuracy stands as a cornerstone for investors and analysts alike. Existing methods still struggle with the complex dynamics of markets, especially in adapting to stock fluctuations' multifaceted nature. In response to these limitations, the proposed work introduces a ground-breaking approach that integrates the robust forecasting capabilities of VARMAx with the adaptive ability of Deep Dyna Q algorithms. The rationale for this integration is rooted in the quest to enhance the responsiveness and accuracy of stock value predictions. The model's performance was evaluated using datasets from the Indian and USA markets. It showed significant improvements in precision, accuracy, recall, AUC, and specificity, as well as a substantial reduction in response time. These enhancements are crucial for financial decision-making, where accuracy and timeliness are essential.

Social media has over the years emerged as a powerful platform for communicating and sharing views, thoughts, and opinions. However, at the same time it is being abused by certain individuals to spread hate against individuals, communities, religions etc. Such content can lead to serious issues of mental health, online well-being, and social order. Therefore, it is very important to have automated methods and approaches for detecting such content from the large volume of posts in social media. Recently there has been several efforts to develop computational approaches towards this end, however, most of these efforts are directed towards content in English language. Only recently studies have started focusing on low resource languages, including those from South Asia. This paper attempts to present a detailed and comprehensive survey of hate speech related research in South Asian languages. The various definitions and terms related to Hate speech in different social media platforms are discussed first. The different tasks in the hate speech research, available datasets and the popular computational approaches used in the South-Asian languages are surveyed in detail. Major patterns identified and the practical implications are presented and discussed, along with a discussion of challenges and opportunities of further research in the area.

The current research aims to optimize the multi-response during Wire Electrical Discharge Machining (Wire-EDM) of SLMed AlSi10Mg by using Taguchi integrated Grey Relational Analysis (GRA). Selective Laser Melting (SLM) is one of the most well-known and practical Additive Manufacturing (AM) methods, with the potential to replace many traditional production processes. The complex metallic support structures created during SLM require special attention as they are difficult to remove by hand. Therefore, this study performs post-processing using the Wire-EDM precision machining technique to evaluate the machinability of the SLMed AlSi10Mg as-built part. The multi-response optimization in this study aims to achieve the maximum material removal rate and the lowest surface roughness while considering four important influencing factors (pulse On time, pulse Off time, servo voltage, and wire feed rate) at four distinct levels. The Taguchi integrated Grey Relational Analysis (GRA) revealed that a pulse On time of 118 μs (Level 3), a pulse Off time of 44 μs (Level 1), a servo voltage of 60 V (Level 4), and a wire feed rate of 7 m s⁻¹ (Level 4) are recommended to achieve optimal machining of SLMed AlSi10Mg. Furthermore, the derived optimization results were carefully verified through a confirmatory experiment, which showed a 38.57% improvement in multi-response characteristics compared to the initial Wire-EDM parameter settings. The methodology proposed in this work offers a standardized approach that has the potential to be implemented for the rapid and precise prediction and optimization of surface roughness, while achieving better material removal during Wire-EDM of SLMed AlSi10Mg.

Aim Satellite images are significantly more accessible to collect and include a huge amount of informative data in selected geographical areas. However, because of their vast dimensions and acquisition procedures, information extraction or segmentation is an extremely complicated procedure. So, this paper proposes a satellite image segmentation technique to extract required information that can be applied to real-time applications. Background Satellite images are vast sources of information that help to perceive the earth’s surface and relevant changes in it. In satellite imaging, image segmentation is crucial as it leads to better classification and understanding of the data present in the considered images. Objective This paper presents an enhanced segmentation technique based on color-based fuzzy cmean clustering (FCM) presents an improved segmentation technique based on color-based fuzzy c-mean clustering. Method One of the popular types of soft clustering techniques that is utilized for image segmentation is fuzzy c-mean clustering. It is chosen for its robust features in data categorization. This study suggests an FCM method for segmenting colored satellite images based on clusters created using the colors red, green, and blue. Result The performance of the proposed system is done with seven test images by comparing the segmented output of each image obtained by the popular threshold technique and the proposed methodology. Four performance metrics are employed in quantitative analysis to assess the effectiveness of the proposed method: entropy, standard deviation (SD), PIQE (Perception Image Quality Evaluator), and NIQE (Naturalness Image Quality Evaluator). A higher value of Entropy denotes better quality of images and a lower value of NIQE shows more intricate image details. In both the parameters, the images obtained by the proposed techniques showed better quality as per an increase in their entropy value and a decrease in NIQE value. Conclusion The traditional threshold-based methods are applied to assess the performance of the proposed methodology utilizing four image-measuring parameters: entropy, NIQE, PIQE, and standard deviation. Overall, better results are obtained in all test cases using the proposed FCMbased clustering technique.Applying qualitative as well as quantitative analysis, the proposed method has compared the performance of the threshold technique with the proposed approach using seven satellite images. Experimental images taken from the public domain.

In USA, out of all the carcinomas, one of the most rampant variety of carcinomas is skin cancer, with an estimated one in five Americans developing it by the age of 70. As per the Skin Cancer Foundation, in the USA alone, every hour more than 2 people succumb to skin cancer. For melanoma skin cancer, the survival rate could be 99% considering a 5-year time frame if it is detected early. Deep learning, a subdomain of AI, empowers computers to learn complex patterns from massive amounts of data. Convolutional neural networks (CNNs), an eminent deep learning architecture, along with its variations like VGG19, MobileNet, ResNet, ResNext, and the latest Vision transformers excel at image recognition tasks, making them ideally suited for analyzing medical images like skin lesions. This review explores the burgeoning utilization of deep learning in skin cancer detection. The analysis of the constraints of conventional methods and highlights of the potential of deep learning in achieving superior accuracy and objectivity have been discussed in this study. The review methodology involves a comprehensive search of relevant research papers and publications from Google Scholar. The review focuses on the studies involving deep learning for classification or segmentation of skin cancer, enabling more efficient and trustworthy AI systems. The findings reveal CNNs as the mainstay, with both traditional training and transfer learning approaches proving effective. However, recent advancements showcase the promise of vision transformers, ensemble methods, and hybrid models, alongside innovative augmentation and optimization techniques, combining attention layers with state-ofthe- art architectures, making clinically trustworthy systems using XAI techniques like GRADCAM, leading to significantly improved efficiency. In conclusion, this review emphasizes the transformative power of deep learning algorithms for the diagnosis of skin cancer, paving the way for highly accurate, trustworthy, and accessible diagnostic tools and presents an analysis of the latest developments related to AI and deep learning architectures and frameworks being applied for diagnosis of skin cancer.

Iris biometrics allow contactless authentication, which makes it widely deployed human recognition mechanisms since the couple of years. Susceptibility of iris identification systems remains a challenging task due to diversity in spoof or presentation attacks (PAs) that fails to assure consistency while adopting them in real life scenarios. Hence, iris PAs are the growing concerns that gained significant attention in recent past decade. To alleviate these attacks or recognize presentation attack instruments (PAIs), iris presentation attacks detection (IPAD) algorithms are designed to distinguish a real and fabricated iris trait. Aiming at the efficient iris spoof detection mechanism, in this research work we expound a novel ensemble learning-enabled model (IensNet) that learns three pre-trained and fined-tuned deep models (i.e. DenseNet161, ResNet and VGGNet) for better accuracy and generalized performance. The novel IensNet approach offers several merits (i.e. consolidated strengths of multiple models, improved generalization ability, etc.) as compared to a simple transfer learning strategy where the knowledge is drawn from single pre-trained model. Finally, our approach learns a novel fully-connected dual layer classifier via outcome of three fine-tuned models to yield a final classification result as bonafide or spoof iris trait. Our approach is evaluated on Notre Dame LivDet iris 2017 and Notre Dame contact lenses 2015 anti-spoofing datasets. The experimental analysis of IensNet offers outstanding performance with a lower ACER of 0.2% and 1.4% for Iris-LivDet-2017 and Notre Dame contact lenses 2015 dataset respectively. Besides, IensNet exhibit promising results in cross-dataset environment with an ACA of 91.46%.

Excessive rainfall and droughts harshly impact India's social and economic growth. Though several statistical methods have been used in literature to predict Indian monsoons, uncertainties cannot be ruled out. The accuracy prediction of ISMR (Indian Summer Monsoon Rainfall) is scientifically demanding. From this perspective, it is essential to explore exploiting machine learning techniques. In this paper, a novel De-correlated Regularized Random Vector Functional Link Neural Network Ensemble (DRRNE) prediction approach was proposed using Climate Predictors such as Southern Oscillation Index (SOI), Sea Surface Temperature Anomaly (SST), El-Niño Southern Oscillation (ENSO), and Dipole Mode Index (DMI) to predict ISMR. The proposed work has also investigated the predictability of climate above predictors using the DRRNE approach to predict ISMR. In addition to the predictors above, the data for an 8-year training window time series for June to September is combined and analyzed for four predictors (ENSO, DMI, SOI, and SST) to derive another predictor, ENSO-DMI-SOI-SST (EDSS). It is found that the combination of these four predictors- the EDSS- produces better accuracy than using any of the individual predictors in this study. Among the individual predictors (ENSO, DMI, SOI, and SST), the DMI predictor has shown the best predictability for ISMR prediction. Thus, the suggested study concludes that the DRRNE technique with negative correlation learning may be a suitable tool for predicting the ISMR using the combined outcome of the four climate predictors as mentioned above.

Sharing information has been one of the main goals for ages. The need for more efficient and interactive sharing methods increased as time passed. This gave birth to the “World Wide Web”, which enabled searching for information amongst different documents. Initially, everything was static, meaning users could only read the information from the web. With the growth of social media, a new Web model emerged: a two-way flow of information. Now, users not only read information from the web but can also share it. The productivity level can be increased tremendously by integrating various technologies like Artificial Intelligence, Virtual Reality, Augmented Reality, Blockchain and Internet of Things (IoT). With these new technologies, a new era of the web is emerging in which, instead of traditional 2D websites, a 3D virtual space will be created where users can meet their friends face-to-face via virtual avatars, create, buy and sell digital artefacts or tokens (NFTs). This chapter aims to cover these new emerging technologies and what difference they can make in lives and the whole working structure of the Internet.

Federated Adversarial Learning (FAL) maintains the decentralization of adversarial training for data-driven innovations while allowing the collaborative training of a common model to protect privacy facilities. Before sharing with bigger global aggregation, it allows users to change settings locally over many iterations. However, a strong network against attackers in Industry 5.0 towards consumer digital ecosystems is a challenge for adversarial training methodologies. To solve this issue, a novel FAL-based Customized Inequality-Aware Federated Learning (CusIAFL) technique is proposed in this paper for classifying and securing color images. The proposed method reduces the instability brought on by the heterogeneity of the data and optimizes each data sample by understanding the client-label distribution. A unique Pix2Pix Generative Adversarial Network (GAN) algorithm is employed to generate realistic images in the presented research work, while a hybrid approach is used to guarantee consistency in the time series data. This innovative research work is evaluated on various non-medical, consumer electronic, and medical imagery. The experimental results have been evaluated using performance metrics, namely accuracy, entropy, Peak Signal-to-Noise Ratio (PSNR), Hausdorff Distance (HD95), Structural Similarity Index (SSIM), and Mean Square Error (MSE). The results show that the proposed technique outperforms the existing models in terms of security.

The demand for real-time, high-quality services (QoS) is increasing with the proliferation of the resource-constrained nature of edge devices that facilitate the Internet of Things (IoT) and wireless sensor network (WSN) applications. Several existing multi-objective algorithms, such as MOPSO, Elitism MOGA, MODE, and others, are capable of balancing exploration and exploitation; they assist in efficient QoS management for WSN-IoT applications, address resource limitations, and align with the objectives of the applications. However, they suffer from showing robustness in the solution and efficient convergence rates on benchmark functions impacting overall QoS. This paper proposes a multi-objective optimization and edge-intelligent adaptation-based strategy to address QoS management issues, jointly optimize several competing objectives, like energy and latency, and maximize localization and coverage rates while considering the limitations of edge devices. The proposed work uses a novel Grey-wolf optimizer (GWO) Algorithm with an innovative bird-edge-computing adaptation approach to analyze the complex connections between input parameters, edge resources, and QoS indicators to generate Pareto-optimal solutions. The evaluation of the proposed edge intelligence technique with IoT applications demonstrates its effectiveness compared to conventional heuristic-based approaches. This approach enhances the QoS in IoT applications and improves resource utilization and scalability in edge computing environments.

Ensuring safe pregnancy and reducing maternal and infant mortality rates require early prediction of fetal health. The application of machine learning algorithms in monitoring fetal health helps to improve the chances of timely intervention and better outcomes in case of any possible health issues in fetuses. Existing studies offered to aid this issue typically by training models using a significant portion of the dataset, ranging mostly around above 70%. The only existing active learning method in this field employs around 41% training samples to achieve 98% accuracy. This work presents a novel active learning technique to identify the most informative data samples for training a model leading to high accuracy with limited number of training samples. It employs a novel query function built upon uncertainty and diversity criteria which are derived based on properties of XGBoost classifier and distance from each other. For deriving uncertainty criterion the soft probabilities obtained for the unlabelled samples are used while the distance among the uncertain samples in feature space is utilized for deriving diversity criterion. The proposed approach shows superior performance in comparison to all the state-of-the-art methods. Through analysis and experimentation, the proposed solution achieves an accuracy above 99% by utilizing less than 20% of the dataset for training. This demonstrates its efficacy and potential in fetal health monitoring. The code and dataset is available at the GitHub repository https://github.com/niyg7/fetal-health-dataset.

Heart disease is a worldwide health concern for which precise risk assessment and early detection need a call for solutions that are creative as well as accurate. Cardiovascular research has undergone a significant revolution because of advancements in computational intelligence (CI) techniques like machine learning (ML), which has improved diagnostic accuracy and identified new risk factors. To predict the risk of heart disease in the early stages, ML algorithms evaluate large chunks of diversified patient data, while also considering their lifestyle, genetic markers, and medical history. Some of the meticulous features for careful engineering and selecting methods required to create effective ML models include feature extraction, dimensionality reduction, hyperparameterization, etc. The decision support systems often provide pragmatic insights suitable to individualized treatment suggestions. These features of ML-based heart disease prediction are a beacon to bridge the gap between these predictions and actual clinical practices. Therefore, it would be suitable to conclude that ML has great potential to address patient-specific therapies, the early diagnosis of the disease, and the risk assessment in the context of heart diseases. This paper compares the performance of various CI approaches in heart disease prediction. It evaluates the performance of different evaluation metrics by varying the train test splits. It will help the researchers working in the relevant domain to select the most suitable model for designing the heart disease diagnostic system.