Volume 13 | Issue 5 |

 Abstract

Agriculture is very important for food supply and the economy. However, plant diseases are a big problem for farmers because they can reduce the amount and quality of crops. Finding and treating plant diseases early is important to avoid major losses. Normally, farmers or experts check plants by looking at them directly, but this method is slow, hard work, and sometimes incorrect, especially when farming on a large scale. To solve this problem, we created a system that uses Convolutional Neural Networks (CNNs), a type of deep learning model that is very good at understanding images. Our model can look at pictures of plant leaves and detect if the plant is healthy or has a disease. It also tells the user the name of the disease and suggests how to treat and prevent it. We built a web application using Flask, where users can easily upload a leaf image. The CNN model quickly analyzes the image and gives accurate results. We used the PlantVillage dataset with over 54,000 images of different plant diseases to train the model. Our model achieved about 95% accuracy, showing it can detect plant diseases very well. This system can help farmers make faster and better decisions about their crops, reduce the use of chemicals, and improve overall farm productivity. In the future, we plan to add more types of plants, create a mobile app for easy use in the field, and connect the system with IoT devices for automatic monitoring. Overall, this project shows that AI can make farming smarter, more efficient, and more sustainable.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Plant Disease Detection, Deep Learning, Convolutional Neural Networks (CNN), Image Classification, PlantVillage Dataset, Web Application.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

As technology has developed and sensors have been miniaturized; there have been attempts to employ contemporarytechnologyinavarietyoffieldstoimprovethe qualityofhumanlife.Oneoftheimportantareasofresearch that has been identified is the inclusion of technology in the healthcare business. People who require healthcare services find them unreasonably expensive, especially in developing countries. As a result, this effort aims to solve a current healthcare concern in society. The primary purpose of the projectwastodeveloparemotehealthcaresystem.It'sbroken intothreeparts.Thefirstcomponententailsusingsensorsto detect a patient's vitals, the second sending data to cloud storage,andthethirddeliveringtheobserveddataforremote viewing.Thedatamaybeseenremotely,allowingadoctoror guardian to monitor a patient's health state even while they arenotinthehospital.TheInternetofThings(IoT)has-been widely used to connect readily available medical resources and give patients with intelligent, dependable, and effective healthcare services. Health monitoring for active and supported living is one of the paradigms that may use IoT benefits to improve the patient's lifestyle.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Patient Monitoring, IoT, Smart Device, Smart Health Monitoring, Medical Services.

  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

WiMAX, OFDM, Wireless Communication, AWGN, Interleaving, Trellis Coding

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study focuses on developing an automated garlic price prediction system in India using a hybrid model that integrates Random Forest and Long Short-Term Memory (LSTM) networks. The system leverages historical price trends, and market conditions to enhance prediction accuracy. Individual models, including ARIMA, SVR, Prophet, XGBoost, and LSTM, were evaluated, with XGBoost demonstrating the highest standalone performance. However, due to the limitations of single models in capturing both non-linear relationships and sequential dependencies, a hybrid approach combining Random Forest for feature selection and LSTM for time-series learning was adopted.Visualizations, including geospatial mapping and district-wise price distribution, provide insights into price variations across different regions.Future enhancements may include incorporating external socio-economic factors, real-time data integration, and advanced deep learning techniques to further improve forecasting accuracy. This research contributes to agricultural price intelligence systems, potentially benefiting farmers, policymakers, and market analysts in optimizing market strategies and mitigating price risks.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Garlic Price Prediction, Machine Learning, LSTM, Mean Square Error, ARIMA, SVR, Prophet, XGBoost, Hybrid Model(Random Forest + Long Short Term Memory)

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Effective recruitment techniques are crucial for organizational success in the service-based sector, which depends heavily on human resources. The impact of employer branding on hiring practices in the service industry is examined in this study. It looks into how people's opinions about a company's image, values, and work culture affect hiring decisions and draw in top talent. The study examines how important aspects of corporate branding, including social media presence, organizational reputation, and employee value proposition, influence candidate behaviour using both primary and secondary data

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Employer branding, Organizational reputation, Recruitment, Competitive employment market.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The Normalized Difference Vegetation Index (NDVI) is widely used for analyzing vegetation health through satellite imagery. It is computed as the normalized difference between red and near-infrared (NIR) spectral bands, allowing per-pixel greenness estimation. NDVI leverages multi-spectral remote sensing data to classify land cover, identify vegetation density, and distinguish features such as water bodies, barren lands, and forested regions. This study presents a web-based NDVI analysis framework that integrates Sentinel Hub for satellite image acquisition, Streamlit for interactive visualization, and SQLite for structured report storage. The system enables users to dynamically retrieve Sentinel-2 imagery, compute NDVI, EVI, and SAVI in real-time, and visualize results across multiple spectral bands (True Color, Infrared, and Vegetation Indices). A dynamic interactive map powered by OpenStreetMap allows users to intuitively select geographic locations for analysis, enhancing accessibility and user experience. Additionally, the system provides histograms, surface plots, and statistical insights, allowing users to analyze vegetation patterns efficiently. The system also supports data export in CSV format and structured local storage for historical analysis. By automating NDVI computation and visualization, this research enhances the efficiency of environmental monitoring, reducing the manual effort associated with traditional GIS-based approaches. Future enhancements include cloud deployment for real-time large-scale analysis, integration with a more robust database system, and development of a full-stack web application for expanded functionality and scalability.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Remote Sensing, NDVI Sentinel Hub, SQLite, Streamlit, Vegetation Index, OpenStreetMap.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The usage of mobile phones has increased significantly all over the world as a result of the introduction of the internet and the passage of time. When it comes to mobile operating systems, Android is the one that is used on the most devices. There has been a significant increase in the amount of malicious applications that are designed to target mobile devices as a result of the widespread availability of smartphones. The discovery of new forms of malicious software and versions of harmful software has gotten increasingly difficult in recent years. Some Keywords for Our Method Through the use of correlation distance, it is possible to correlate important codes that are contained within Android malware source code. API calls, Android permissions, common arguments, and phrases that are regularly used are all included in these crucial codes. As a result, the Support Vector Machine (SVM) is utilized, which makes it possible to recognize both newly developed malicious software as well as malware samples that have already been constructed. In contrast to more conventional ways, this strategy is distinguished by the fact that it is not associated with the context of the text. The actions of dangerous programs are documented through the use of this approach, which involves the integration of an operating system with many forms of malicious software.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Android, malware, the correlation distance between keywords, and the super vector machine (SVM).

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Social media has become a significant platform for communication, especially among the youth. Unfortunately, the rise of cyber bullying on social networking sites has raised concerns. To address this issue, researchers have explored the use of machine learning to automatically detect social media bullying actions, focusing on identifying patterns and similarities in the language used by bullies. While various textual-based techniques have been employed for social media bullying detection, our model aims to showcase the implementation of software designed to detect bullied content, such as tweets and posts. This project introduces an intelligent machine learning-based system to detect abusive tweets in real-time using logistic regression and TF-IDF vectorization. Drawing parallels from IoT-based cable fault detection , our system identifies the 'fault' in language content and alerts users immediately. A Streamlit-based web interface demonstrates real-time detection of abusive tweets.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Abusive tweets, machine learning, logistic regression, cyberbullying , NLP, TF-IDF, streamlit

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Background: Urinary tract infections (UTIs) are among the most common infections affecting women of all age groups. Lack of knowledge regarding preventive measures leads to increased frequency, recurrence, and complications. Aim: This study was conducted to assess the level of knowledge regarding urinary tract infections and their preventive strategies among women in selected areas of Mohali, Punjab. Methodology: A descriptive research design was adopted with a purposive sampling technique. A structured questionnaire was administered to 60 women to assess their knowledge about UTI causes, symptoms, and prevention. Data was analyzed using descriptive and inferential statistics. Results: Among the 60 participants, 26.7% had good knowledge, 53.3% had average knowledge, and 20% had poor knowledge. Significant associations were found between knowledge levels, and educational status as well as previous history of UTI. Conclusion: The study revealed moderate knowledge among most women, but gaps still exist. Health education interventions are essential to reduce incidence, recurrence, and improve preventive behavior.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Urinary Tract Infection, Women's Health, UTI Prevention, Knowledge Assessment, Hygiene Practices, Public Health.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The study focuses on the employee health and safety at KAUSIKH THERAEUTICS (P) LTD. its critical role in enhancing organizational productivity and employee well-being. The study explores various workplace hazards, including physical, chemical, biological, and ergonomic risks, and evaluates the effectiveness of existing health and safety policies in mitigating these risks. The sample size is 217. Data was collected through surveys, interviews, and workplace observations across multiple industries, providing valuable insights into employee awareness, safety training practices, and the implementation of safety standards. The findings indicate a strong correlation between proactive safety measures and reduced workplace accidents, as well as improved employee morale and performance. The study emphasizes the importance of regular risk assessments, employee participation in safety programs, and continuous training to foster a culture of safety. Ultimately, the research underscores that investing in employee health and safety is not only a legal and ethical responsibility but also a strategic approach to achieving long-term organizational success.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Health and safety.

  License

Creative Commons Attribution 4.0 and The Open Definition