Volume 14 | Issue 7 |

 Abstract

Online platforms now distribute news at a pace that far exceeds the capacity of manual verification. Fake news written in a formal, journalistic register consistently evades single-model classifiers and leaves users without any explanation for an automated verdict. This paper presents NewsZ, a hybrid explainable AI platform for multi-source fake news detection that unifies four complementary techniques: an ensemble of classical machine learning classifiers (Naive Bayes, Logistic Regression, Support Vector Machine, Random Forest), a two-layer Long Short-Term Memory (LSTM) network, claim-level fact-checking against indexed fact-check databases, and a SHAP-based explainability module that produces keyword evidence and natural-language narratives.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Fake News Detection; Hybrid AI; LSTM; Explainability; SHAP; Fact-Checking; Web Content Extraction; Uncertainty Quantification; Ensemble Learning; Misinformation

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This research proposes an AI-based smart medical dispensing system for secure and real-time access to medicines in rural and semi-urban areas. The system includes prescription scanning, Optical Character Recognition (OCR), and machine learning based validation to accurately identify medicines and enforce controlled dispensing. The model extracts medicine details from scanned prescriptions to ensure safe and regulated distribution and prevent misuse. A precision-controlled dispensing mechanism enables accurate single-tablet dispensing for optimal dosage safety and reliability. The system is equipped with secure multi-mode payment options and a solar-powered design with battery backup for 24/7 uninterrupted operation. The proposed system enhances accessibility, safety and efficiency of healthcare, contributing to the potential of AI-based automated kiosks in advancing smart healthcare solutions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Smart medical dispensing, Rural healthcare, Prescription validation, AI based automation, Controlled drug delivery, OCR based scanning, Secure healthcare systems

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Road accidents resulting in loss of life continue to pose significant challenges globally. Delayed emergency response due to inadequate information about the accident scene aggravates the situation further. Therefore, the paper proposes an Internet of Things (IoT)-based vehicular accident detection and reportage system that seeks to mitigate the aforementioned issues.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

IoT, Accident Detection, ESP32, GPS, GSM, ThingSpeak, Vehicular Data Logging, Road Safety, Embedded Systems.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Online learning environments often lack effective mechanisms to monitor student engagement and emotional states during lectures, which can negatively impact learning outcomes. To address this challenge, we propose an adaptive multimodal fusion framework that integrates facial expression analysis and speech-based emotion recognition for real-time engagement estimation. The video branch employs a Separable 3D Convolution (S3D) model to capture spatio-temporal facial features, while the audio branch utilizes a CNN-BiLSTM architecture to model temporal speech patterns. Predictions from both modalities are aligned to engagement levels and fused through an adaptive weighting mechanism that dynamically adjusts based on individual confidence scores. Experimental evaluation will be conducted on benchmark datasets such as DAiSEE and RAVDESS, demonstrating that multimodal fusion significantly outperforms unimodal approaches. The proposed system enhances robustness and reliability across diverse e-learning scenarios, supporting intelligent environments that adapt instructional strategies to improve student engagement and overall learning effectiveness

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Multimodal fusion, student engagement detection, facial expression analysis, speech emotion recognition, e-learning systems.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Waste segregation is a critical challenge in modern urban environments, especially in countries like India where large volumes of mixed waste are generated daily. Traditional waste segregation methods heavily rely on manual work, making the process time-consuming, error-prone, and unsafe. To address this, we propose a system with a hybrid pipeline using YOLOv8m and EfficientNet for detection, classification, and intelligent decision-making. The system collects real-time video input, converts it into frames, and preprocesses images for detection. YOLOv8m detects waste with 64% accuracy across 5 classes(plastic, metal, glass, paper, organic) , EfficientNet verifies it with 94% accuracy across 8 classes(plastic, metal, glass, paper, cardboard, organic, textiles, wood), and the data is passed to a secondary decision engine to further classify waste as biodegradable, recyclable, or hazardous. The system handles multiple objects under real-world conditions and helps analyze waste trends, improves recycling efficiency, and reduces human effort, making it suitable for real-time deployment.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Waste Segregation, YOLOv8, EfficientNet, Deep Learning, Computer Vision, Smart Waste Management

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The recruitment and placement process in educational organizations most of the times are separated, slow, and they don't have standard methods for evaluating candidates. Students, colleges, and employers each work from their own separate platforms, and this causes a lack of coordination, less opportunities for interview practice, and different methods of assessing candidates. In order to solve these problems, this article suggests an AI-based integrated platform that brings all the parties together into one single system characterization. Artificial intelligence is used by the system so that it can simulate the interviews in real-time and also carry out automatic evaluation of the candidate based on their technical knowledge, communication skills, and confidence.It is developed with an expandable MERN-P architecture, which includes React. js for the frontend, FastAPI for the backend asynchronous processing, MongoDB for the data storage with high flexibility, and Supabase for the authentication and management of structured data. The platform has separate portals for the different categories of users like students, companies, and even administrators which allow them to interact with each other without any hindrance, manage the data in a centralized manner and keep track of the performance.Automated feedback and analytics are two of the great tools that can be used to boost student preparedness as well as the decision-making process. The solution put forward by this paper is one that will indeed bring in the factor of rise in productivity, giving room for growth and ensuring that the evaluation is carried out from the standpoint of fairness, thereby making this system the right fit for the contemporary recruitment.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Artificial Intelligence, Recruitment System, Mock Interview FastAPI MERN Stack NLP,Real-Time,Systems,StudentPortal,Placement System.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

India has emerged as one of the world's fastest-growing major economies, demonstrating remarkable resilience despite global economic disruptions, geopolitical tensions, and post-pandemic challenges. The vision of transforming India into a five trillion-dollar economy represents not merely an economic milestone but a comprehensive developmental agenda aimed at improving living standards, fostering innovation, strengthening infrastructure, promoting inclusive growth, and enhancing global competitiveness. Achieving this ambitious target requires sustained high economic growth, structural reforms, technological advancement, increased productivity, and effective governance. This study examines India's progress toward the five trillion-dollar economy objective by analysing the key growth drivers, sectoral performance, policy initiatives, and emerging opportunities. It further explores the major constraints, including employment generation, infrastructure gaps, income inequality, regional disparities, fiscal pressures, and global economic uncertainties. The paper adopts a descriptive and analytical research approach using secondary data collected from government publications, international organisations, research institutions, and scholarly literature. The findings suggest that India's demographic dividend, expanding digital economy, vibrant startup ecosystem, manufacturing initiatives, infrastructure development, and growing domestic market provide a strong foundation for long-term economic expansion. However, sustaining high growth will require continuous investment in human capital, skill development, technological innovation, green infrastructure, financial inclusion, ease of doing business, and institutional reforms. The study concludes that a balanced policy framework combining economic efficiency with social equity and environmental sustainability is essential for transforming India's long-term economic aspirations into reality.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Five Trillion-Dollar Economy, India, Sustainable Growth, Economic Development, Infrastructure, Digital Economy, Manufacturing, Employment, Public Policy, Inclusive Growth.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Banana wine is an alcoholic beverage prepared through the fermentation of banana juice using yeast. Bananas are rich in carbohydrates, vitamins, minerals, and natural sugars, making them suitable raw materials for wine production. The present study focuses on the preparation and fermentation of banana wine under controlled laboratory conditions. Fresh ripe bananas were processed to obtain juice, followed by the addition of sugar and yeast for fermentation. The fermentation process was carried out for several days at room temperature. Physicochemical parameters such as pH, alcohol content, titratable acidity, aroma, color, and taste were analyzed periodically. The results indicated successful fermentation with desirable sensory characteristics and acceptable alcohol production. Banana wine can be considered a nutritious and economically valuable fermented beverage with commercial potential.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Banana wine, Fermentation, Yeast, Alcohol production, Fruit wine, Physicochemical analysis.

  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

Fermentation , Yeast Growth, Alcohol Estimation, Physico-Chemical Analysis, Fermentation Efficiency.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Eco packaging from pomegranate is a better replacing for the conventional plastic packaging properties. Biodegradable and active food packaging materials derived from pomegranate peel extract ,biopolymer. The present study aimed to develop a biodegradable food packaging film incorporated with pomegranate peel extract and evaluate its antimicrobial properties against selected foodborne pathogens. Pomegranate peel extract was prepared using solvent extraction and incorporated into a starch-based biodegradable film using the casting method. The antimicrobial activity of the developed films was evaluated against common foodborne bacteria including Escherichia coli and Staphylococcus aureus using the agar diffusion technique. The results indicated that films containing pomegranate peel extract inhibits the growth of notable microorganisms. activity compared with control films without extract. The inhibition effect increased with higher concentrations of the extract. These findings suggest that biodegradable films enriched with pomegranate peel extract may serve as promising eco-friendly packaging materials capable of enhancing food safety and extending shelf life of the products.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Biodegradable, Eco packaging, Biopolymer, Food borne, Agar diffusion, Shelf life.

  License

Creative Commons Attribution 4.0 and The Open Definition