Volume 13 | Issue 5 |

 Abstract

Machine learning classifiers were evaluated for their effectiveness in detecting fraudulent ad clicks. Among the models tested, the Random Forest algorithm demonstrated the strongest performance. Incorporating predictive models not only improved the click-through rate but also enabled the system to learn from data autonomously, eliminating the need for manual domain expertise. However, a significant challenge in this field is the limited availability of well-labeled, publicly accessible datasets, which restricts the scope of experimentation and analysis. Fraudulent ad clicks can be artificially generated and mixed with genuine user activity on websites, complicating detection efforts. In a comparative analysis, the Support Vector Machine (SVM) achieved the highest standalone accuracy at 93.3%, while other machine learning models ranged between 84% and 92%. In ensemble approaches, combining XGBoost with Random Forest also yielded a 93.3% accuracy rate, whereas the combination of SVM and XGBoost resulted in a lower accuracy of 76.6%. Overall, both SVM and the ensemble of XGBoost and Random Forest demonstrated the best predictive performance.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Advertise, fraud click, Ensemble machine learning, prediction, user, legitimate, link

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Human activities that harm biodiversity have reached increasingly higher thresholds in the modern era. Human actions that significantly negatively influence ecosystems and biodiversity include forestation, overfishing, pollution, and climate change. These activities have resulted in a reduction in the diversity of life on Earth. Biodiversity provides enormous ecological, social, and economic benefits and is critical for maintaining natural balance. To counterbalance the negative impacts of human activity on biodiversity, conservation efforts are required to protect species and their habitats, conserve genetic variety, and support the sustainable use of natural resources. Conservation operations can be carried out in various ways, including protected areas, environmentally sound farming and fishing methods, and the execution of environmental regulations. The consequences of this biodiversity decline are far-reaching impact, including ecosystem resilience, resource availability, and the ability to respond to environmental change. The severity and root causes of habitat degradation and biodiversity loss must be determined. Furthermore, priority areas for conservation efforts may be identified, and the establishment of environmental regulations must be considered. The main purpose of this research paper is to determine human activities; impact on biodiversity and the necessary conservation strategies.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Deforestation, Climate Change, Ecosystem, Biodiversity, Environment.

  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

Synthesis and Characterization of Magnesium nano ferrite

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

AgroVolt X is a solar-powered irrigation system designed for rural farming. It offers a sustainable solution by utilizing renewable energy and smart sensors to optimize water use. This affordable, eco-friendly technology enhances crop productivity, reduces energy costs, and supports self-reliant, environmentally sustainable agriculture in remote and underserved areas.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

AgroVolt X, renewable energy, solar-powered irrigation, rural farming, sustainable agriculture, smart irrigation, water conservation, eco-friendly technology, energy-efficient farming, rural development.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

In the last few years, several techniques for facial manipulation in videos have been successfully developed and made available to the masses (i.e., Face Swap, deepfake, etc.). These methods enable anyone to easily edit faces in video sequences with incredibly realistic results and a very little effort. Despite the usefulness of these tools in many fields, if used maliciously, they can have a significantly bad impact on society (e.g., fake news spreading, cyber bullying through fake revenge porn). Even though the technology was developed by experts, now it is available as web application and mobile application which enables normal people to access it and make manipulated videos and photos easily. Then it has become a big threat to the society. The ability of objectively detecting whether a face has been manipulated in a video sequence is then a task of utmost importance. One of the points to select this topic as the project is this relevance of this problem in the society. This project is intended to detect deepfake manipulation in videos. The proposed method is based on the concept of ensembling. Indeed, it is well-known that model ensembling may lead to better prediction performance. The models that are used for the prediction are DenseNet and GoogleNet. DenseNet (Densely Connected Convolutional Networks) consists of Dense blocks. In the blocks, the layers are densely connected. GoogleNet architecture is based on the notion of having filters with multiple sizes, which can process on same level. Also, this is a comparative study of the performance of both the models.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

DenseNet, GooleNet, CNN, GAN,Deepfake

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Karl Marx's critique of the class-based school system is rooted in his broader analysis of capitalist societies, where institutions, including education, serve to perpetuate existing class structures. Marx viewed education not as a neutral or liberating force but as a mechanism that reinforces social hierarchies and sustains the dominance of the ruling class. Marx's analysis of the class-based school system reveals how education can function as a mechanism for maintaining social hierarchies. By reproducing class inequalities, legitimizing them through the myth of meritocracy, and conditioning individuals to accept their social roles, the education system plays a pivotal role in sustaining the capitalist order.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Revolutionizing , Meritocracy , superstructure, Capitalist , Relevance , Pedagogy.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This paper deals with practicing more than one religion by a person named Pi Patel in Ang Lee's film Life of Pi. Pi Patel explores Hinduism, Christianity and Islam during his young age. He is a Hindu by birth and practices all the three religions despite of confusion. Practicing more than one religion is not a bed of roses. Religion is part and parcel of every body's life. It has very important place in one's life especially in India where multi-religions are in practice. The film is based on Yann Martel's novel of the same name. The film shows richness of religious belief and strong faith in God in the term of practicing three different religions by a young boy.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Religion, Film, Novel, Visual, Globally, Philosophical, Journey.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Research is the backbone of any country. The world's most powerful nations have the best research and development culture. Research in recent years has yielded results that were never thought of before, who would have thought a decade ago about 5G data at such nominal prices in India, thanks to research? Obviously, the answer is 'Yes'. In contemporary times, with the introduction of NEP 2020, UGC has introduced various initiatives for the betterment of the quality of research done by research scholars. UGC chairman Prof. Mamidala Jagadesh Kumar, once pointed out that most of the research theses uploaded on Shodhganga are substandard. This raises a big question about scholars' knowledge concerning the process of research and research analysis. The study is concerned with the 6th theme of the National Seminar, i.e. "Data-Driven Decision Making in Education", and more precisely the sub-theme 6.3 "Educational Research and Big Data Analytics". The study was conducted to check the Achievement in Research of Scholars as it is an important part of educational research & data analytics, taking into consideration, the basics of research and data analysis. The objective of the Study was to study the influence of Gender, Educational Qualification, NET Qualification and their various interaction on Achievement in Research of Scholars. The hypothesis was that "There is no significant influence of Gender, Educational Qualification, NET Qualification and their various interactions on Achievement in Research of Scholars". The sample consisted of 100 Scholars, belonging to Post Graduate (P.G.) and Ph.D. courses of the State Universities of Uttar Pradesh, selected through the Random Sampling Technique. Survey Research Method was used by the researchers. The tool that was used by the researcher for data collection was an unstandardised self-made test to test the Achievement in Research of Scholars, which was administered to the scholars either via a Google Form or via a hard copy. The data were collected in the Excel sheet and then analysed with the help of the 'Three Way ANOVA' statistical technique. The Achievement in Research was found to be dependent on Gender, NET qualification, interaction between Gender and Educational Qualification, and interaction among Gender, Educational Qualification and NET Qualification.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Research, Achievement, Scholars, Three Way ANOVA, State Universities and Uttar Pradesh.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Depression continues to be a critical concern in the landscape of global mental health, particularly as individuals increasingly express their psychological states through digital communication. This study proposes a comprehensive sentiment analysis framework that leverages Natural Language Processing (NLP) and Machine Learning (ML) methodologies to identify signs of depression from textual input. The system incorporates sequential stages including text normalization, tokenization, lemmatization, and TF-IDF-based feature extraction. Multiple machine learning algorithms were assessed on a multiclass mental health dataset. While XGBoost demonstrated high training accuracy, it was prone to overfitting. Logistic Regression was selected as the final model due to its more stable and consistent performance on unseen data. Additionally, the framework integrates probabilistic outputs and word cloud visualizations to enhance interpretability, making it suitable for scalable and noninvasive depression detection applications.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Depression Detection, Sentiment Analysis, Natural Language Processing, Machine Learning, Text Mining, Logistic Regression

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This paper introduces a solar PV-powered Dynamic Voltage Restorer (DVR) utilizing a Trans-Z-source Inverter (Trans ZSI) to improve power quality in grid-connected photovoltaic (PV) systems. The DVR, a power electronic device, compensates for voltage disturbances by injecting the required voltage at the Point of Common Coupling (PCC). Compared to conventional Voltage Source Inverters (VSI), the Trans ZSI provides enhanced buck/boost capabilities, a broader voltage gain range, fewer passive components, and lower voltage stress, making it a more efficient alternative for DVR applications. To enhance disturbance detection, ensure precise voltage compensation, and minimize harmonic distortion in the injected voltage, a hybrid control strategy combining the Unit Vector Template and Maximum Constant Boost Control (UVT-MCBC) is proposed. The effectiveness of the Trans ZSI-DVR is analyzed under various voltage disturbances, including swell, interruptions, minor sag with harmonics, and severe sag conditions. Comparative simulations and performance analysis highlight the superiority of the proposed system over traditional ZSI-DVR and VSI-DVR configurations. The results indicate significant improvements in mitigating voltage sag, swell, and interruptions while enhancing the power quality of both the PCC voltage and the PV system's output.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

PV, DVR, Trans ZSI, voltage sag, voltage transient, THD, Unit Vector Template (UVT), Harmonic Distortion

  License

Creative Commons Attribution 4.0 and The Open Definition