Volume 14 | Issue 4 |

 Abstract

Situation awareness (SA) is a crucial factor affecting flight safety for pilots, yet few studies have focused specifically on modeling SA for pilots, resulting in limited success. In this paper, we propose a novel multimodal deep learning approach to monitor pilots' SA. The approach combines handcrafted and deep features obtained from eye movement and flight control data collected from 27 novice pilots across different training phases using a flight simulator. Ground truth SA measurements were obtained using the Situation Awareness Global Assessment Technique (SAGAT). The handcrafted features included 13 eye movements and 22 flight control features, while deep features were extracted from time-series of gaze positions using a deep extractor based on Transformer. By fusing the handcrafted features of eye movement and flight control, along with one deep feature of eye movement, we predicted the final SA level. Through leave-one-flight-out cross-validation, our model achieved a higher accuracy of 92.04%. The results indicate that the multimodal model outperforms the unimodal models, with the eye movement modality demonstrating superiority over the flight control modality in predicting SA. This suggests our method provides an objective means of predicting pilot's SA and offers new insights for SA assessment in aviation and other fields.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Human factors in aviation, pilot situation awareness, virtual profile for pilot

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

DrugInsight predicts drug-drug interactions (DDIs), returning an interaction probability, a severity tier, and a structured clinical explanation for any queried drug pair. The system combines AttentiveFP graph neural network encodings of molecular structure with curated pharmacological data from DrugBank and post-market adverse event signals from TWO SIDES. RDKit and PyTorch Geometric convert SMILES strings into molecular graphs; the GNN encodes each graph into a 256-dimensional structural embedding. These embeddings are concatenated with a 12-dimensional feature vector derived from shared biological targets, metabolizing enzymes, transporters, and carriers (DrugBank), and the Proportional Reporting Ratio from TWOSIDES. An MLP classifier processes the combined vector. A tiered fusion layer then weights the DrugBank rule score, MLP output, and TWOSIDES signal according to what evidence is available for the queried pair, producing a 0-100 risk index and a severity label. We evaluate the model under a drug-level cold-start split: neither drug in any validation pair appears during training. This tests whether the system can handle newly approved or understudied compounds rather than simply interpolating over seen chemical space. The best checkpoint reaches a validation AUC of 0.7065. Clinical explanations are generated by a rule based module covering CYP enzyme-mediated metabolic competition, shared pharmacological targets, and pharmacovigilance signal strength; no language model is involved, so explanation outputs carry no hallucination risk. Per-source confidence scores accompany every prediction. The system is accessible via a Streamlit web interface, a FastAPI endpoint, and a command-line interface.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Drug-drug interactions, graph neural networks, AttentiveFP, evidence fusion, explainability, pharmacovigilance, DrugBank, TWOSIDES

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Nowadays e-commerce platforms are growing rapidly; a number of people visit websites and also use them. On the other side data, privacy and security challenges also increase. If these privacy and security challenges are not resolved, consumers can never trust or visit any e-commerce websites. This study provides a comparative analysis of the data security and data privacy in traditional and modern digital system. It gives a detailed evaluation of different aspects of data privacy and issues related to its implementation. It also examines the need to secure the personal data of the users and the frequent security concerns that online business encounter. This paper also examines strategies that are helpful for users to stay safe in the digital era by taking security precautions. The results have shown that the efficiency of modern security solutions, their scalability and ability to detect threats real-time are better than the traditional approaches, but modern solutions have also brought privacy and data transparency challenges. Recent cybersecurity statistics establish the validity of the analysis and provide an understanding of the need to develop advanced security frameworks to tackle the rapidly increasing cyber threats.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Data security, E- commerce, E- commerce privacy

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Millions worldwide suffer from chronic pain, which impairs quality of life and function. Pharmaceuticals and surgeries can cause dependency, side effects, and limited long-term efficacy. Naturopathy and yoga are holistic, non-pharmacological, and patient-centered alternatives to chronic pain management. Naturopathy emphasizes the body's natural ability to heal through diet, herbal remedies, hydrotherapy, acupuncture, and physical therapy. Through asanas, pranayama, and meditation, yoga can reduce chronic pain by improving flexibility, stress, and neurobiological changes that increase pain perception. Both methods address physical, psychological, and physiological pain to reduce arthritis, fibromyalgia, lower back pain, and migraines, according to research. Scientific validation, regulatory standardization, and integration into traditional healthcare remain despite growing evidence supporting these techniques. This study examines the efficacy, mechanisms, and potential integration of naturopathic and yogic therapies for chronic pain management, emphasizing personalized, evidence-based approaches. To improve the credibility and accessibility of complementary therapies for chronic pain, clinical trials, mechanistic studies, and policy reforms are needed.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Chronic Pain, Naturopathy, Yoga, non-pharmacological, physical therapy, clinical trials.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This project proposes a government-driven sustainable tourism platform for Kerala that leverages advanced AI capabilities to enhance travel planning, monitor visitor experiences, and generate measurable sustainability insights. The system integrates a Gemini-powered intelligent engine to provide real-time itinerary recommendations tailored to user preferences, environmental conditions, and local regulations. By dynamically optimizing travel routes and activities, the platform promotes eco-friendly tourism while improving overall visitor satisfaction.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Keywords-- Sustainable Tourism, Kerala Government, Responsible Tourism Mission, Google Gemini API, NLP, Flask, React, JWT Authentication, PDF Report Generation

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The rapid evolution of cyber threats such as Distributed Denial of Service (DoS), backdoor intrusions, and advanced persistent attacks has made traditional Intrusion Detection Systems (IDS) increasingly ineffective. Signature-based systems fail to detect unknown or zero-day attacks, leading to significant security vulnerabilities. This paper presents a Deep Learning-Based Network Intrusion Detection System (DL-NIDS) designed for real-time monitoring and classification of network traffic. The proposed system utilizes a two-stage deep learning architecture trained on the UNSW-NB15 dataset to first distinguish between normal and malicious traffic, and then classify the attack into specific categories such as DoS, Backdoor, and Generic attacks. The system integrates real-time traffic monitoring, data preprocessing, machine learning inference, and an interactive web-based dashboard for visualization. Experimental results demonstrate improved detection accuracy, reduced false positives, and efficient real-time performance compared to traditional IDS approaches.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Intrusion Detection System, Deep Learning, Network Security, Real-Time Monitoring, Cyber Threat Detection, UNSW-NB15

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

India is secular country and source of knowledge , Values, Norms of society should have the ground of this i.e. secularism. And Religion Has revealed it self as Ideology system of thoughts located in the domain of sacred. But at present India we can observe that, Religion continuously has been customising for Build a new form of that by the force of politics. I named it Religious modernism. That's mean it is a New Replica of the old Version of Architects. And what is the impact of this on society, is it all over negative weather positive, I discussed in my paper. Even not only that, I also discussed about how India and his neighbour countries especially Bangladesh and Pakistan are Constantly doing Support and Spread Religious Intolerance in world. And is Religious Tolerances that things that we need in this Situation? In India we can see the Many kind of impact of the Religious Modernism and Political Exclusivism . And not only that at the present time it is being caused of Mass-hysteria . So for this reason I also discussed what is centre, ground, and future of this social phenomena. And which sphere of society is mostly targeted and influenced by this. But in spite of this there are Few atheists and I also tried to lightened the roll of them in a democratic world. In this present world we can see that Religion is being driven by politics and politics is being Navigated by Religion. This Customised form of religion can become a blessing for any country, and what is the present religious Scenario in India I also discussed about it. And at last I have tried to Make it clear that can we say that Being Spiritual is identical with being religious? how India is shifting torwards a new Religious order where religion is beyond of Spiritual property.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Religious modernism, Secularism, Intolerance, Acceptance , Mass hysteria, Politics Religion, Pseudo Religion.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study examines the financial literacy of tribal and non-tribal households in two villages of the Ghatsila Block, Jharkhand, using data collected from 240 households. A Financial Literacy Index (FLI) was developed following the OECD/INFE framework, which captures financial knowledge, behaviour, and attitudes. The findings show clear differences in financial capability across household groups, with tribal households generally displaying weaker financial understanding and more limited engagement with formal financial practices. Analysis of the three dimensions of financial literacy reveals significant gaps in basic financial knowledge, moderate performance in day-to-day financial behaviour, and mixed attitudes toward long-term financial planning. The logit model further indicates that socio-economic factors such as education, gender, employment type, and access to bank accounts influence the likelihood of being financially literate. These patterns highlight the need for context-specific financial education initiatives tailored to rural and tribal communities.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Financial literacy, Tribal households, OECD framework, Rural Jharkhand, Financial inclusion.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The competitive job market presents significant challenges for candidates, where approximately 40% of applicants are rejected due to performance anxiety and a lack of confidence. Traditional preparation methods, such as static guides or informal practice, often fail to provide the objective, micro-level feedback required for corporate success. This research proposes a new approach using a unified, serverless ecosystem to enhance interview readiness through an AI-powered Resume Analyzer and a Real-Time Voice-Agent. By integrating the Google Gemini 2.0 API for semantic parsing and the Vapi Web SDK for low-latency WebRTC-based audio orchestration, the system extracts critical professional metadata and evaluates delivery fluency via prosodic feature analysis. System evaluation demonstrated a highly effective Semantic Parsing Accuracy (SPA) of 94.2% for the ATS module and an average Round-Trip Latency (RTL) of 1.15 seconds for the real-time voice agent. The aim of this research is to propose a data-driven pipeline that democratizes professional coaching and eliminates visual demographic biases to ensure ethical governance and inclusive assessment by relying strictly on voice and text analytical modules.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Large Language Models, WebRTC, Semantic Parsing, Real-Time Audio, Prosodic Analysis, Mock Interview System, Applicant Tracking System (ATS), Bias Mitigation, Generative AI

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Typical live USB systems do not retain runtime changes, causing all configurations and data modifications to be discarded after shutdown. This limitation restricts their use in scenarios requiring continuity and repeatability. To address this issue, a compact persistence-enabled USB solution is proposed, integrating automated disk initialization, Debian system deployment, and snapshot-based state management. The storage medium is configured using a GPT scheme, where a Btrfs partition functions as the primary writable layer. This partition is internally divided into dedicated subvolumes for the operating system, user data, and snapshot storage. Data compression using zstd is applied to reduce write overhead and improve flash storage efficiency. Boot support is implemented for both UEFI and legacy BIOS platforms, ensuring compatibility across diverse hardware. A snapshot management mechanism records system states and dynamically incorporates them into the GRUB boot menu, enabling direct restoration of previous configurations without requiring external tools or network connectivity. Experimental validation across multiple systems confirms stable boot behavior and consistent preservation of system state across power cycles, demonstrating the effectiveness of the proposed approach for portable and reliable computing environments.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Persistent Bootable USB, Btrfs-Based Storage Design, Portable Debian Environment, Hybrid Firmware Boot (UEFI/BIOS), Filesystem Snapshot Management, Automated OS Deployment, GRUB Boot Integration, Flash Storage Efficiency, Hardware-Independent Computing, Offline System Restoration

  License

Creative Commons Attribution 4.0 and The Open Definition