Volume 14 | Issue 5 |

 Abstract

The present study titled "A Study on the Awareness and Preference of Investors Towards Digital Gold Investment in Coimbatore" aims to examine the level of awareness and perception of investors regarding digital gold as an emerging investment option. With the rapid digitalization of financial services, investment avenues have evolved significantly, offering more convenience and accessibility. Digital gold allows investors to purchase gold in small quantities online and store it securely without physical handling. The study is based on both primary and secondary data. Primary data was collected through a structured questionnaire distributed among investors in Coimbatore, while secondary data was gathered from journals, websites, and financial reports. The study analyses factors such as awareness level, investment behaviour, preference over traditional gold, satisfaction level, and problems faced by investors. The findings indicate that while a significant number of investors are aware of digital gold, there is still a need to increase awareness regarding its safety, reliability, and long-term benefits. The study concludes that digital gold has strong potential for growth due to its convenience, affordability, and security, but greater efforts are needed to build trust among investors.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Digital Gold, Investment, Investor Awareness, Preference, Financial Technology

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Generative artificial intelligence now produces architectural imagery faster than the discipline can evaluate it. As AI-generated images enter concept-stage architectural workflows, the key question is no longer whether AI can generate compelling visuals, but how such images should be judged as architectural propositions. This paper argues that the necessary criterion is fidelity to design intent: the degree to which an AI-generated image remains aligned with the experiential, cultural, material, and temporal intentions that informed it. The paper proposes a conceptual-methodological five-stage AI-assisted pipeline for evaluating fidelity to empathetic design intent in AI-generated architectural imagery. Structured around the TACT framework - Tactility, Atmosphere, Culture, and Temporal Identity - the method embeds evaluation across the full workflow: material ingestion, TACT-informed analysis, prompt construction, image generation, comparative assessment, and human validation. The framework is illustrated through the Ethiopian New International Airport concept, where the design intent centers on Ethiopian identity, highland light, Rift Valley spatiality, tactile materiality, threshold hospitality, and temporal continuity. The contribution is methodological: a structured, auditable, and bias-aware framework for evaluating whether AI-generated architectural imagery remains faithful to empathetic architectural intent.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Generative AI; empathetic architecture; TACT framework; fidelity to design intent; AI-assisted design evaluation; architectural image judgment; cultural specificity; temporal identity.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The purpose of this systematic literature review is to examine existing research on effect of Social media and internet on health of students of higher education institutions. The study identifies relevant articles published between 2015 and 2025 in various databases. To encourage responsible use of internet and social media in education, the review emphasises the need for increased public education, convenient and accessible use of internet and social media and various options, for the same and the development of effective communication strategies. The study also identifies gaps in the literature, such as a lack of research on specific populations and the need for additional research into the effectiveness of interventions aimed at ensuring proper and healthy use of internet and social media in education. Overall, this review offers insights into the current state of knowledge on effect of Social media and internet on health of students of higher education institutions and areas for future research and interventions to encourage responsible internet use.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

social media, effect of social media and internet, health of students of higher education institutions

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

HealthGuard AI is an IoT based healthcare monitoring and emergency detection system developed using ESP32 microcontroller technology. The proposed system continuously monitors important health parameters such as heart rate, body temperature, body movement direction, and fall detection using integrated sensors including pulse sensor, DS18B20 temperature sensor, and MPU6050 accelerometer sensor. The collected sensor data is processed in real time and displayed on an OLED display module. The system also provides wireless monitoring through a web dashboard using WiFi connectivity. During emergency conditions such as abnormal body temperature or sudden fall detection, the system automatically activates a buzzer alarm and sends emergency notifications through Telegram API. This research work aims to improve patient safety, remote monitoring capability, elderly care support, and healthcare accessibility using low-cost IoT technology.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

ESP32, IoT, Healthcare Monitoring, Fall Detection, Pulse Sensor, Telegram Alert, Smart Healthcare, Embedded System

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract Endometriosis is an estrogen-dependent inflammatory condition that affects approximately 10% of reproductive age women worldwide, or an estimated 190 million women. Although it has a high clinical and socioeconomic burden, the average latency to diagnosis is 7-11 years, which is intolerable for patients. The structural normalisation of severe pain during menstruation, a lack of provider awareness and a historical clinical preoccupation with laparoscopy are mostly responsible for this diagnostic delay. Failure to detect or misdiagnose this condition is compounded by extreme symptom variation and by a large overlap between gastrointestinal, urological and musculoskeletal disorders, such as irritable bowel syndrome, interstitial cystitis and low back pain. To overcome these systemic bottlenecks, and to overcome the data fragmentation of many different clinical repositories scattered around, the field is transitioning to artificial intelligence (AI) and machine learning (ML) architectures. This systematic review assesses comprehensively the use of computational systems that decode multi-modal, high-dimensional biomedical data streams for improved risk stratification and diagnostic optimisation. Supervised learning classification models, like AdaBoost, Random Forest and XGBoost algorithms coupled with unique mobile health screening applications have shown outstanding predictive capacity reaching area under the receiver operating characteristic curve (AUC) as high as 0.94 using patient reported symptom patterns. At the same time, natural language processing (NLP) systems such as the context-aware transformer models (e.g., BERT) and longitudinal generative forecasting models (e.g., Foresight) are achieving unprecedented accuracy in extracting comprehensive phenotypic information, discovering complex networks of comorbidities, and identifying trends of misdiagnoses in the unstructured passages of the electronic health record (EHR) text. Further, innovative machine learning techniques are successfully leveraging non-invasive diagnostics and optimizing the automated analysis and categorization of complex biomarker panels (circulating microRNAs and endometrial BCL-6 expression), and improving pelvic imaging segmentation and accuracy on transvaginal ultrasound and magnetic resonance imaging systems, thereby eliminating unnecessary pre-surgical risks. Lastly, this analysis examines the use of explainable AI tools, like LIME and SHAP, to demystify the inner workings of complicated algorithms and foster vital physician trust that helps remove the "black box" hurdle. Finally, it is crucial to address the variations in datasets from inconsistent patient recruitment mechanisms and achieve universal semantic interoperability across different, global, and independent systems of terminology, such as the ICD and SNOMED-CT common ontologies, to enable equitable, objective, and person-centred gynecological care across the globe.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Endometriosis, Diagnostic Delay, Machine Learning, Natural Language Processing, Electronic Health Records, Multimodal Biomedical AI.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Phishing attacks persist as one of the most financially destructive forms of cybercrime, deceiving users through fraudulent emails, malicious URLs, and spoofed web pages to harvest sensitive credentials and payment data. Conventional blacklist and rule-based defenses are fundamentally inadequate against newly registered phishing domains and zero-day threats that evade static catalogues. Machine learning has emerged as a powerful adaptive alternative, enabling detection systems to learn discriminative patterns from labeled data and generalize to previously unseen attacks. This paper presents a structured comparative review of machine learning and deep learning methods for phishing detection, synthesizing findings from fourteen recent studies spanning classical classifiers, ensemble methods, deep neural architectures, and reinforcement learning. Algorithms examined include Random Forest, SVM, XGBoost, LightGBM, Artificial Neural Networks, CNN, RNN-GRU, and hybrid ensemble models. Comparative analysis reveals that multi-modal ensemble and sequence-based deep learning approaches achieve the highest detection accuracy, with reported values reaching 99.98% for email phishing and 98.74% for large-scale URL classification. Persistent challenges including dataset imbalance, zero-day evasion, and real-time latency constraints are critically discussed. Future directions encompassing explainable AI, federated learning, transformer-based detection, and lightweight on-device models are proposed.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Phishing Detection, Machine Learning, Deep Learning, URL Classification, Random Forest, Convolutional Neural Network, Cybersecurity

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

There is increasing awareness regarding health issues in recent times due to health issues related to lifestyle, which are being faced by people based on their eating and living habits. This trend is resulting in increasing attention being paid to systems that can assist people in making tailored suggestions regarding food and drinks for them rather than general suggestions. There has been significant research done with the help of computing techniques like Artificial Intelligence, Machine Learning, Deep Learning, Retrieval-Augmented Generation, and nutritional information analysis in order to improve recommendation accuracy. However, despite such advancements, there remain issues related to usability factors such as health-based filtration, ingredient control, and simple user interface. In order to overcome such challenges, this research proposes a Smart Juice Recommendation System that caters to people who need to be aware of their diets and be healthy as well. The system takes into consideration inputs relating to the physical state, lifestyle, health aspects, and ingredients availability of the users in order to come up with appropriate drink recommendations. Unlike giving out general output, the system provides personalized recommendations that suit user nutritional suitability. For generating such decisions, the system employs simple rule-based algorithms and BMI calculation and nutrition filters . Functionality of the system is enhanced by incorporating ingredient tracking, nutritional display, personalization feature, history retention, notification capability, and voice-enabled output. A lightweight implementation approach using Python with user-friendly GUI is adopted for this purpose. The development of the entire solution has been driven by prior research on healthcare recommendation systems and nutrition-enabled computational algorithms . In the end, the system helps the users make better decisions regarding their choice of drinks according to their health needs. It shows how simple computer-based methods can help in personalizing nutrition. Possible improvements may include incorporating machine learning, storing data on the cloud, connecting with wearable devices, and developing mobile applications.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Personalized Nutrition, Juice Recommendation System, BMI Calculation, Health Management System, Python Based Application, Nutritional Profiling, Artificial Intelligence.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Indan's rising electricity demand and regional disparities require state-level forecasting for effective energy planning. This study presents a cluster analytics-based framework to forecast electricity supply positions and peak gap indicators across Indian states using historical data on energy availability, demand, installed capacity, transmission constraints, and consumption trends. Cluster analysis and statistical forecasting estimate future demand, availability, and surplus or deficit conditions. Results show inter-state differences in electricity adequacy, highlighting the need for infrastructure investment, renewable integration, and demand-side management.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Electricity forecasting, Cluster analytics, Peak demand gap, Indian power sector, Energy planning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The skill of metal hand engraving of names on utensils is one of the oldest, sustainable and most enduring forms of artistic expression. The legacy of writing names and dates on new utensils has been one of the significant and lifelong traditions while gifting, which reflects memories and remembrance of emotional context in the relationships. Such skill to maintain the legacy of relationships and emotional connection is endangered due to the challenges in maintaining skilled artisans, lack of diversification, upgradation and more over unexplored out of its utensil segment in terms of design application. To address this challenge, utilizing the same skill of engraving out of its typical usage is the need of the hour. Hence, transforming this functional skill into creating aesthetically pleasing products is the focus of the research. The objective of the paper is to revitalize the unexplored application of hand engraving skill onto an aesthetic & commercial product. The research not only uplifts the skill of the artisans but also converts the same into a promising handcrafted product segment with an amplified value to cherish. Keeping in mind the original essence of hand engraving on utensils, a well thought product range of jewellery is created by converting miniature utensils with extraordinarily hand engraved surfaces. The exploratory research incorporates primary and secondary data collection through interviews and questionnaires with diminishing artisans from the local markets in Pune city and craft lovers, followed by an apt design process to develop products. This study delves into the intricacies of the skill, into a craft that explores a modern re-emergence in the form of unique jewellery pieces for women that are innovative and sustainable. The research has not only contributed to reviving the uniqueness of the unexplored skill but has also contributed towards promotion, sustenance and economic empowerment of the diminishing artisan's community.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Metal Hand Engraving, Sustainable Craft Practices, Miniature Utensils, Artisan Empowerment.

  License

Creative Commons Attribution 4.0 and The Open Definition