In the rapidly evolving field of cybersecurity, effective threat behaviour identification is essential for identifying and mitigating threats, particularly in a virtualized environment. In this paper, we propose a hybrid VMM-IDS security framework, called DeepIntrospector, that leverages both network and system artifacts of applications to enhance threat detection capabilities. In our method, we extract critical artifacts from both network traffic and system memory logs of emerging network malware families at the hypervisor level. The proposed traffic behaviour monitoring mechanism develops a deep learning model using bidirectional long short-term memory(Bi-LSTM) to identify the malicious network behaviour from traffic logs. The proposed system behaviour monitoring mechanism develops a deep learning model using a dense neural network (DNN) from system logs. Each of the model assigns a prediction scores to the executables, indicating their potential threat level. By applying the alert fusion mechanism, we synthesize these scores to enhance the accuracy of the prediction of the proposed security model. The framework is validated using the emerging malware dataset, and the results seem promising.

Network Intrusion Detection Systems (NIDS) are essential for securing modern network infrastructures, yet detecting emerging attacks remains challenging due to dynamic traffic behavior, class ambiguity, and mislabeled traffic flows. To address these issues, this work proposes an explainable and efficient NIDS framework, termed X-NetIntrospector. The framework performs network introspection in a virtualization environment to capture malware traffic flows, followed by a Gaussian-distribution-based filtering mechanism to remove ambiguous traffic flows. A Kullback–Leibler (KL) divergence-based feature selection strategy is then employed to improve class-wise discriminative learning while reducing computational complexity, followed by deep learning-based training. To enhance performance, a counterfactual explanation (CFE)-driven post-training label correction mechanism is introduced to identify and rectify mislabeled flows. Model interpretability is further supported by peer-loss analysis and explainable AI (XAI) techniques. The proposed framework is validated on a self-generated dataset (NetDB) and the NSL-KDD dataset, demonstrating improved detection performance under naturally occurring label imperfections. In addition, performance is evaluated under explicit noise-injection scenarios, in which a preliminary label-refinement stage enhances data consistency prior to learning, complementing the post-training correction strategy of the core detection pipeline. Experiments on the CICIDS-2017 dataset under symmetric and asymmetric explicit noise-injection settings provide promising results.

Virtualization can be defined as the backbone of cloud computing services, which has gathered significant attention from organizations and users. Due to the increasing number of cyberattacks, virtualization security has become a crucial area of study. In this paper, we propose an explainable and introspection-based malware detection approach called vDefender for fine-grain monitoring of virtual machine (VM) processes at the hypervisor to identify the malicious behaviour of 17 different malware families of Windows exhibiting new evolving behaviour. Initially, it performs a basic security check to detect hidden processes and ensures the presence of security-critical processes. Then, deep memory introspection is performed using a software breakpoints injection approach to intercept the execution of processes. Various process activity logs are captured that include process-related, file manipulation, kernel heap object creation, exception-related activities, etc. Hybrid feature vectors are derived from these logs, which are reconstructed using the proposed mechanism to eliminate the redundant behaviour. The features are then learnt using Random Forest (RF) algorithm to classify distinct malware families. The interpretation and analysis of RF results involve the use of explainability techniques. The proposed approach achieves an accuracy of 95.49%, F1-score of 95.82% with 0.05% false alarms when evaluated using an emerging malware dataset. The contribution includes a comprehensive discussion of results, accompanied by a comparative analysis of current approaches that gives readers insight towards future research directions.

In the era of digitalization, electronic gadgets such as Google Translate, Siri, and Alexa have at least one characteristic: They are all the products of natural language processing (NLP). “Natural Language” refers to a human language used for daily communication, such as English, Hindi, Bengali, etc. Natural languages, as opposed to artificial languages such as computer languages and mathematical nomenclature, have evolved as they have been transmitted from generation to generation and are challenging to explain with clear limits in the first instance. In natural language processing, artificial intelligence (Singh et al., 2021), linguistics, information processing, and cognitive science are all related fields (NLP). NLP aims to use intelligent computer techniques to process human language. However, NLP technologies such as voice recognition, language comprehension, and machine translation exist. With such limited obvious exclusions, machine learning algorithms in NLP sometimes lacked sufficient capacity to consume massive amounts of training data. In addition, the algorithms, techniques, and infrastructural facilities lack enough strength. Humans design features in traditional machine learning, and feature engineering is a limitation that requires significant human expertise. Simultaneously, the accompanying superficial algorithms lack depiction capability and, as a result, the ability to generate layers of duplicatable concepts that would naturally separate intricate aspects in forming visible linguistic data. Deep learning overcomes the challenges mentioned earlier by using deep, layered modelling architectures, often using neural networks and the corresponding full-stack learning methods. Deep learning has recently enhanced natural language processing by using artificial neural networks based on biological brain systems and Backpropagation. Deep learning approaches that use several processing layers to develop hierarchy data representations have produced cutting-edge results in various areas. This chapter introduces natural language processing (NLP) as an AI component. The history of NLP is next. Distributed language representations are the core of NLP's profound learning revolution. After the survey, the boundaries of deep learning for NLP are investigated. The paper proposes five NLP scientific fields.

A field experiment was conducted during rainy (kharif) season of 2016 at Hazaribag, Jharkhand to study the effect of varieties and phosphorus (P) applications on grain yield and economics of direct-seeded upland rice (Oryza sativa L.). The experiment was laid out in split plot design with three replications, keeping 3 rice varieties, viz. ‘Vandana’, ‘Anjali’ and ‘CR Dhan 40’, in the main plots and 7 different P rates. viz., 0 kg/ha, phosphorus solubilizing bacteria (PSB) + arbuscular mycorrhizal fungi (AMF), 13.2 kg P/ha through single super phosphate (SSP), 13.2 kg P/ha through SSP + PSB + AMF, 26.4 kg P/ha through SSP, 26.4 kg P/ha through SSP + PSB + AMF and 39.6 kg P/ha through SSP in the sub-plots. The results revealed that yield attributes, viz. effective tillers/m2 (312), filled grains/panicle (130), panicle weight (27.9 g), 1,000-grain weight (22.5 g), and grain (3.91 t/ha) and straw yields (6.17 t/ha) were higher in ‘CR Dhan 40’ compared to ‘Anjali’ (259, 115, 2.04 g, 20.4 g, 3.33 t/ha and 5.64 t/ ha respectively) and ‘Vandana’ (252, 108, 1.78 g, 19.9 g, 3.12 t/ha and 5.53 t/ha respectively). Similarly, application of 39.6 kg P/ha through SSP recorded the highest values for yield attributes, grain (4.25 t/ha) and straw yields (6.79 t/ha), and was at par with 26.4 kg P/ha through SSP + PSB + AMF. Net returns and B:C ratio were also recorded the highest in ‘CR Dhan 40’ with 39.6 kg P/ha through SSP, which was at par with 26.4 kg P/ha through SSP + PSB + AMF. Thus, integrated application of 26.4 kg P/ha through SSP along with PSB and AMF in ‘CR Dhan 40’ can enhance grain yield and net returns of direct-seeded upland rice in Jharkhand.

Artificial intelligence (AI) and machine learning have emerged as very promising technological advancements in recent times, exhibiting extensive potential for application across a wide range of industries, including the healthcare sector. Cancer is a highly frequent non-communicable disease that is a leading cause of mortality on a global scale. Scientists have conducted extensive investigations in order to enhance the lethality and invasiveness of cancer. The application of artificial intelligence in cancer research has been extensively employed, yielding highly promising results thus far. Various strategies can substantially enhance the prognosis of individuals with cancer, with particular emphasis placed on timely detection and accurate diagnosis facilitated by a range of imaging modalities and scientific methodologies. One of the myriad applications of artificial intelligence (AI) in the field of medical research is to its utilization as a method for enhanced detection and diagnosis. The primary objective of this study article is to comprehensively examine the existing literature and provide a comprehensive overview of the various applications of artificial intelligence (AI) in different commonly occurring cancers. Age-related skeletal disorders, such as cancer, infection, and osteoporosis, provide a substantial challenge within contemporary societies. Although there are existing professional interventions available for the treatment of these illnesses, it is important to note that several of these interventions include significant hazards. The presence of various pathogenic mutations and the aggregation of hereditary illnesses can contribute to the development of cancer and an elevated mortality rate. The proliferation of malignant cells, which can manifest in any bodily organ or tissue, poses a significant risk to an individual’s overall well-being. Cancer, sometimes referred to as a tumor, necessitates accurate and expeditious early identification in order to identify viable therapeutic options. Bone cancer is a matter of considerable medical importance due to its frequent association with patient mortality. The utilization of pictures obtained from X-ray, MRI, or CT scans is employed in the diagnosis of bone malignancies. Osteosarcoma is a neoplastic condition characterized by the presence of a malignant tumor, typically occurring in the long bones of the limbs. The increasing incidence of cancer and the imperative for healthcare services have rendered the task of identifying and classifying this ailment more complex. Bone malignancy is an atypical pathological condition characterized by uncontrolled cellular proliferation within the skeletal system. The destruction of bone tissue that is in a state of good health occurs. A bone affected by malignancy will exhibit distinct tactile characteristics compared to an unaffected bone. The collection exhibits morphological similarities between multiple cancerous and healthy bone photographs. Hence, the classification of these entities poses a significant issue. To initiate the process of finding a resolution, we commence by identifying the most efficient method for edge detection and afterward proceed to its construction. Machine learning algorithms are employed to assess the effectiveness of these sets of features.

Stock prediction is a challenging and complicated procedure due to the stock market's great volatility and unpredictability caused by the continual changes in stock prices. When making purchases or sales, traders and investors in the financial markets rely on stock research and trading. In order to get an edge in the markets, traders and investors use data analysis, both historical and current, to make educated judgements. Stock market prediction is and has always been an important field of study for traders and financiers. A stock market prediction is an effort to foretell the future value of an exchange-traded financial instrument, such as a company's shares (Nifty & Sensex), or any other similar asset. When it comes to explaining how stocks are anticipated, our technique makes use of machine learning, which employs a range of models to make prediction more accurate and easy. Stock price forecasting using RNNs equipped with Long Short-Term Memory (LSTM) is the primary research topic here. Our results will be more precise when compared to the current methods used to forecast stock prices. An asset for those who invest in the stock market, a thorough analysis of the company's performance will lead to significant financial gains and useful advice for dealing with problems.