Volume 14 | Issue 4

 Abstract

The present study titled "A Comparative Study of Consumer Preference Between Rapido and Uber Bike Services" aims to analyze and compare the preferences of consumers towards app-based bike taxi services. With the rapid growth of urbanization and increasing traffic congestion, bike taxi services have emerged as a convenient, time-saving, and cost-effective mode of transportation. This study focuses on identifying the key factors influencing consumer choice, such as pricing, safety, availability, service quality, and customer satisfaction. The research is based on both primary and secondary data. Primary data was collected through a structured questionnaire from a sample of respondents, while secondary data was gathered from various websites, articles, and journals. Statistical tools such as percentage analysis, ranking analysis, and chi-square test were used for data analysis. The findings of the study reveal that Rapido is widely preferred due to its affordability and quick availability, whereas Uber bike services are chosen for their reliability, brand image, and safety measures. Factors like convenience, time-saving, and promotional offers also play a significant role in influencing consumer preference. The study concludes that both services have their own advantages, and improving safety features, service quality, and customer support can enhance user satisfaction. The research provides valuable insights into consumer behavior and highlights the growing importance of bike taxi services in modern urban transportation.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Bike taxies services , bike taxi , rapido , uber , bike taxi services

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

What happens when a person is judged before they speak? How does it feel to be denied water? How does it feel to be addressed by something you may not like? These questions are lived experiences and realities for millions of people in India. The caste system decides who a person is, what they will do, where they will live and how other people will treat them. The caste system divides people in society. These are the situations one cannot choose. It is an accident from birth and continues for life. Caste is not just a category; it decides who is powerful and who is not. It divides people into two categories: on top, there are upper-caste people, who get access to education, land, and money. Whereas other people at the bottom, there are castes who are treated as "impure" or "untouchables" also addressed as Dalits, who have faced discrimination and oppression for their entire life.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Caste and Oppression: Voices from the Margins in the Indian Literature

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This paper presents an industrial sorting robot based on ESP32-CAM vision processing and a pick-and-place robotic arm for automated industrial object handling. The system captures real-time video using ESP32-CAM and processes the frames using Python OpenCV for shape recognition. Objects are classified as circle, triangle, square, and rectangle using contour approximation and polygon analysis. Based on the detected shape, serial commands are transmitted to an Arduino-controlled six-servo robotic arm. The robot performs smooth pick-and-place movement using relay-based gripper control and places objects into their corresponding bins. The proposed system significantly improves sorting speed, reduces human intervention, and provides reliable industrial automation performance.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Industrial Automation, Pick and Place Robot, ESP32-CAM, Shape Detection, OpenCV, Arduino, Smart Sorting

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Electricity theft continues to be one of the most per- sistent and economically damaging problems faced by power dis- tribution utilities across the world. Whether it happens through direct meter tampering, bypassing of metering equipment, or falsification of consumption records, the impact ripples across the entire grid - increasing operational costs, destabilizing energy distribution, and ultimately burdening honest consumers with inflated tariffs. Conventional detection methods, which largely depend on manual field inspections or simple threshold-based flagging, are too slow and too imprecise to keep pace with the scale of modern smart meter deployments.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Electricity Theft Detection, Isolation Forest, Anomaly Detection, Smart Meter, Adaptive Retraining, Sliding Window, Machine Learning, Django, Time-Series Classification, Non-Technical Loss

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Cultural heritage is constituted by the dynamic relationship between tangible and intangible dimensions. Historic buildings traditionally functioned as socio-cultural systems where architectural spaces enabled ritual, performance, and collective memory. However, contemporary transformations, functional incompatibility, and deterioration have disrupted this interdependence, leading to cultural discontinuity. This research investigates the relationship between architectural spaces, cultural practices, and community interactions in historic buildings through cultural mapping. By identifying and documenting the interdependencies between space, practice, and users, the study develops an analytical understanding of how built heritage supports cultural activities. The approach shifts the focus of conservation studies from viewing historic buildings as static physical objects to understanding them as living cultural system.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Cultural mapping, cultural practices, historic buildings, intangible heritage, tangible heritage

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Multiplication is a fundamental arithmetic operation that significantly influences the performance of modern digital systems such as processors, digital signal processing units, and embedded systems. Conventional multiplier architectures, including array and sequential multipliers, suffer from limitations such as high propagation delay, increased power consumption, and poor scalability for higher bit-width operations. This project presents the design and implementation of an efficient parallel multiplier architecture to improve multiplication speed and reduce computational delay. The proposed system generates partial products simultaneously and accumulates them using optimized adder structures. Modified Booth Encoding is employed to reduce the number of partial products, while carry save adders are used to minimize carry propagation during partial product accumulation. A high-speed final adder produces the accurate multiplication result. The proposed architecture supports both signed and unsigned multiplication using a unified design, thereby reducing hardware redundancy and improving flexibility. The design is described using hardware description language and verified through simulation and synthesis. The results demonstrate improved performance compared to conventional multiplier designs, making the proposed parallel multiplier suitable for modern VLSI and high-speed digital applications.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

FPGA, Xlinx ISE, Modalism, Matrix Multiplication, Parallel Block-Scheduling.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This research study is to find correlations between gaming disorder, self esteem and academic performances among the high school students of Kannur District. Most of the research existing today is in the context of internet addiction, smart phone addiction, decline in self - esteem and academic performance. But none of the studies were conducted in Indian Territory. The behaviour and addiction status of children are crucial as it brings in lot of challenges both for them as well as to the society as it triggers even the suicidal tendency. The proposal tries to establish correlation by considering a sample of 300 high school students (age 13 - 15), with a positive trend towards gaming disorder. These students will be administered with the shortet version of Gaming Addiction Scale and Coopersmith Self - esteem inventory. Also the academic performance of the subjects will be collected. The data will be analysed using Pearson's correlation coefficient(r), t tests and ANOVA.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Gaming disorder, Self-esteem, Academic performance, High school students, Kannur District, Gaming Addiction Scale, Coopersmith Self-esteem Inventory, Internet addiction, Adolescent behaviour, Pearson correlation, t-test, ANOVA.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Vitamin and mineral deficiencies remain a widespread public health concern, particularly in developing regions where access to diagnostic healthcare is limited. Visible symptoms of such deficiencies frequently appear on external body parts including the skin, nails, eyes, lips, tongue, and hair, making image-based detection a practical and non-invasive screening approach. This paper presents a deep learning-based web application designed to detect vitamin and mineral de-ficiencies from images of human body parts. The proposed system employs InceptionV3, a convolutional neural network pretrained on the ImageNet dataset, fine-tuned through transfer learning to classify six categories of deficiencies: Vitamin A, Vitamin B complex, Vitamin C, Vitamin D, Vitamin KE, and Mineral deficiencies including zinc, iron, biotin, and protein. The dataset used for training and evaluation is publicly avail-able at https://www.kaggle.com/datasets/udaykarthik21bce9252/ vitamin-defficiency-dataset. The model achieves a classification accuracy of 85%. The system is integrated into a Django-based web application supporting user authentication, real-time image-based prediction, confidence score display, and downloadable health reports. This work establishes the feasibility of combining computer vision with accessible web technologies to support early health awareness in a user-friendly manner.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Vitamin Deficiency Detection, Deep Learning, In-ceptionV3, Transfer Learning, Django, Medical Image Classifi-cation, Convolutional Neural Network.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract Unintentional medicine overdose and prescriptionrelated toxicity have become critical public health challenges, particularly in settings with high polypharmacy and limited real-time clinical decision support. Traditional risk assessment techniques rely on manual chart review and static rules, which struggle to capture complex, evolving prescription patterns. In this paper, a machine-learningdriven framework is presented for predicting patient-specific overdose risk using routinely collected clinical and prescription data. The proposed approach utilizes supervised learning models, including Logistic Regression, Random Forest, and Gradient Boosting, to learn patterns associated with high-risk dosage combinations, comorbidities, and prior adverse events. The system is organized as a modular architecture consisting of a data preprocessing pipeline, a model training and evaluation core, and a risk scoring service that can be integrated into clinical applications. Experimental design and evaluation metrics are described to provide a reusable blueprint for academic and project implementations. The results from a prototype implementation indicate that the proposed system can achieve competitive accuracy and recall, demonstrating its potential to support early intervention and safer prescribing practices.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Medicine Overdose Prediction, Machine Learning, Clinical Decision Support, Risk Scoring, Electronic Health Records.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Pesticides are widely used in agriculture and domestic environments, but their improper use poses significant health hazards. Housewives are particularly vulnerable due to their involvement in food handling and household pest control. This study aims to assess the knowledge regarding the effects of pesticides and the protective measures adopted by housewives in selected areas of Nashik, Maharashtra. A quantitative, non-experimental descriptive research design was used. A total of 100 housewives were selected using non-probability convenience sampling. Data were collected using a structured questionnaire. The findings revealed that the majority of participants had moderate knowledge regarding pesticide effects, while a smaller proportion demonstrated good knowledge. Although many participants practiced basic protective measures such as washing vegetables, the use of advanced protective practices was limited. A significant association was found between knowledge levels and selected demographic variables. The study concludes that there is a need for targeted educational interventions to improve awareness and promote safe practices related to pesticide use.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Pesticides, Knowledge, Housewives, Protective Measures, Health Effects, Awareness

  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

bone fracture detection, deep learning, ResNet50, transfer learning, Grad-CAM, musculoskeletal radiograph, DI- COM, explainable AI, web-based clinical tool, two-stage classifi- cation.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Digital technology is slowly changing the way of teaching and learning in the classroom. In many educational institutions, Information and Communication Technology (ICT) is already used to support the learning process. Artificial Intelligence (AI) combined with ICT tools makes the learning process more flexible and effective. This paper examines the role of AI-assisted technologies in supporting skill development within ICT integrated classrooms. Specifically, The discussion focuses on the Indian educational context. A variety of learning activities can be observed using AI-based tools. Based on this information, these systems should provide holistic advice on education and provide feedback in time. Hence, students receive support according to their individual learning needs. The paper also discusses major Indian digital learning initiatives such as DIKSHA, SWAYAM, and PM e-VIDYA. These programmes show the potential of digital platforms to enhance access to educational resources for teachers and students nationwide. At the same time, certain challenges still remain. Issues i.e. limited digital infrastructure, lack of teacher training, and the need for responsible use of AI must be handled carefully. The study indicates that artificial intelligence has the potential to enhance skill-based learning. This occurs when it is integrated with ICT by proper planning and support systems.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Artificial Intelligence (AI), ICT Integrated Classroom, Skill Development, Digital Learning, DIKSHA, SWAYAM.

  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

Anti-dandruff, Carpaine, papaya leaves

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The rapid advancement of Artificial Intelligence (AI) has revolutionized modern healthcare, with particular impact on medical imaging and radiological diagnostics. This paper presents the AI Radiology Co-Pilot, a comprehensive intelligent system developed to assist radiologists in the detection of chest X-ray abnormalities and the automated generation of structured diagnostic reports. The system employs a ResNet-50-based Convolutional Neural Network for accurate image classification, achieving robust detection of pathological conditions including pneumonia, effusion, and cardiomegaly. To facilitate structured report generation, the system integrates Mistral-7B-Instruct, a state-of-the-art Generative AI language model, which converts model predictions into coherent clinical reports and patient-friendly summaries. Additional features include a Grad-CAM-based explainability module for visual interpretation, a multilingual interactive chatbot for patient assistance, and a Flask-based web interface enabling real-time deployment. Experimental evaluation demonstrates significant improvements in diagnostic efficiency, report quality, and clinical communication. The proposed system bridges the gap between image-based classification and language-based report synthesis, offering a unified, interpretable, and accessible AI-powered radiology workflow.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Deep Learning, Medical Imaging, ResNet-50, Generative AI, Mistral-7B-Instruct, Report Generation, Grad-CAM, Chatbot, Healthcare AI, Flask

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Urban civic management systems often suffer from inefficiencies such as lack of transparency, delayed responses, and absence of verification mechanisms in complaint resolution. This paper presents CivicSync, a full-stack web-based e-governance platform designed to bridge the communication gap between citizens and municipal authorities. The system introduces a novel geotagged image verification mechanism that ensures authenticity in both complaint reporting and resolution phases. By integrating browser-based camera access, GPS location tracking, and reverse geocoding using OpenStreetMap services, CivicSync generates tamper-evident visual proof embedded with spatial and temporal metadata. The platform follows a three-tier architecture consisting of citizens, department workers, and administrative authorities, enabling structured workflow management. Additionally, automated email notifications with visual proof enhance user trust and engagement. Experimental implementation demonstrates improved accountability, reduced fraudulent reporting, and enhanced operational efficiency in civic issue management systems

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

( CivicSync, Urban Civic Management, E-Governance Platform, Geotagged Image Verification, GPS Location Tracking, Browser-Based Camera Access, Tamper-Evident Proof, Three-Tier Architecture, Workflow Management, Administrative Authorities, Automated Email Notifications, Operational Efficiency.)

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The incidence of brain strokes has increased, mainly because of lifestyle, health, and delays in early detection. Severe brain strokes result in disabilities or fatalities, thus the need for early diagnosis and prediction. This project presents a creative approach to brain stroke detection and prediction, employing a more accurate and reliable approach through the application of machine learning. The detection model employs deep learning to perform image analysis on CT scans, whereas the prediction model evaluates factors such as age, hypertension, glucose level, BMI, smoking, job type, and living environment to determine stroke risk. This approach takes advantage of existing datasets, applying advanced techniques such as image preprocessing, data augmentation, and transfer learning to boost its performance and reliability The project has also focused on improving the accuracy of the prediction model, ensuring efficient performance with varying datasets. The dual model approach has been adopted to ensure efficient brain stroke detection and prediction, allowing healthcare experts to take preventive measures at an early stage. This will not only reduce brain stroke complications but also contribute to the development of more efficient and intelligent systems

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Machine Learning, Deep Learning, Stroke Prediction, Medical Image Analysis

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Road traffic accidents are a problem that causes deaths and bad injuries all over the world. When Road traffic accidents happen it is really important that people get help away. The old way of watching CCTV cameras is not very good because it mostly relies on people watching them and this can be very slow. The Smart Traffic Accident Detection and Alert System is a way to do things. It uses computer programs to look at CCTV pictures and find Road traffic accidents. This system uses a Convolutional Neural Network to figure out if a Road traffic accident happened. It also uses a YOLO-based object detection model to find the car and draw a box around it. When the Smart Traffic Accident Detection and Alert System thinks a Road traffic accident happened and it is pretty sure it writes down what happened and tells a person to check. The person has 30 seconds to look at it through a web page. If the person does not do anything the Smart Traffic Accident Detection and Alert System will send out emergency messages again and again. The Smart Traffic Accident Detection and Alert System is a way to help people who get hurt in Road traffic accidents. It can find Road traffic accidents quickly. It can get help to people who need it. The Smart Traffic Accident Detection and Alert System is an improvement, over the way of watching CCTV cameras.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Smart traffic systems; accident detection; deep learning; YOLO; computer vision; intelligent transportation systems (ITS); emergency response automation; real-time surveillance; metadata extraction; cloud integration; Flask-based deployment; multi-camera monitoring; decision tagging; analytics dashboard; road safety.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Many small businesses face difficulty in moving to digital platforms. The travel and tourism industry is one of the areas which could benefit from utilizing digital transformation to improve both customer reach as well as the efficiency of providing services. This research paper outlines both the design and development of a new user centered and scalable web application for tour and travel companies. This system will allow small business owners to handle all aspects of their business including managing customer's inquiries and bookings via a single web based interface. The application has been built using HTML, CSS, and JavaScript on the front end and has Firebase for managing authentication (signing up) and databases, while EmailJS has also been implemented to provide email confirmation of bookings and inquiries. With the newly developed application, users have the capability to sign up, sign in, browse through travel packages, and make reservations respectively in a fast and efficient manner. The reservations are stored in an encrypted format and can be retrieved in real time by business owners. This research paper provides a concise and functional application for small business owners that have a formalized structure supporting the digitalization of their operations. As a result, the user experience is enhanced, manual operations are reduced and overall business efficiency is markedly increased. This proposed solution is also scalable and contains the capability of additional modifications such as payment systems and administrative dashboards to be added in the future

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Web Application, Travel Website, Firebase, EmailJS, User Centric Design, Digital Business

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Invasive fungal infections (IFI) are now recognized as a serious global health threats placing a heavy burden on health care system. Treating thus infections is difficult because the available antifungal drugs often cause side effects , have limited effectiveness and face rising resistance. Over the past 30 years,researchers have carried out extensive testing of antifungal drugs.These efforts have been crucial in creating treatment guidelines for serious fungal infections. Most of these guidelines are based on well known and highly cited clinical trials. While we have made progress in developing antifungal medicines that are more effective and less toxic than older options .Early detections and targeted treatment This growing challenge highlights the urgent need for new antifungal medications. This review explores the latest advancements in antifungal drug development, focusing on innovative compounds that are either being tested in clinical trials or are likely to be evaluated soon. These new agents stand out because of their novel mechanisms of action, broader antimicrobial coverage, and improved pharmacokinetic properties. Together, these qualities hold promise for more effective treatment options than traditional drugs.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

: Invasive fungal infections (IFIs), Global health concern, Drug resistance, Clinical trials, Pharmacokinetics, Improved patient outcomes

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The proliferation of deepfake videos synthesized through deep generative models such as Generative Adversarial Networks (GANs) and neural autoencoders poses critical threats to digital security, media integrity, political stability, and public trust. Existing detection approaches either focus exclusively on spatial artifacts within individual frames using CNNs, or on temporal inconsistencies using recurrent networks, but rarely integrate both. This paper presents a novel hybrid deep learning framework combining Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) for robust, video-level deepfake detection. The proposed system employs EfficientNetB0 as the spatial feature extractor to capture per- frame facial artifacts, while a two-layer Bidirectional LSTM models temporal inconsistencies across a sequence of 20 uniformly sampled frames. Face regions are detected using the OpenCV Haar Cascade detector with 25% bounding box padding. The model is trained and evaluated on the FaceForensics++ benchmark comprising 7,000 videos across six manipulation categories: Deepfakes, Face2Face, FaceShifter, FaceSwap, NeuralTextures, and DeepFakeDetection. Class imbalance (1:5 real-to-fake ratio) is addressed through Weighted Random Sampling, label smoothing, and stratified splitting. Training employs AdamW optimization, Cosine Annealing scheduling, and PyTorch Automatic Mixed Precision on an NVIDIA Tesla T4 GPU. Experimental evaluation yields a test accuracy of 92.4%, F1-Score of 91.7%, and AUC- ROC of 0.967, demonstrating the superiority of joint spatial- temporal learning over frame-only and temporal-only approaches. A Flask-based real-time web application is developed for practical deployment, providing binary REAL / FAKE verdicts with confidence scores and REST API integration for downstream systems.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Deepfake Detection, CNN-LSTM, EfficientNetB0, Bidirectional LSTM, FaceForensics++, Video Forensics, Temporal Analysis, Face Manipulation, GAN, Deep Learning, OpenCV, Flask, Real-Time Detection

  License

Creative Commons Attribution 4.0 and The Open Definition