This research paper aims to explore consumer expectations, benefits, and challenges of online shopping for agricultural produce, with the goal of developing and recommending effective marketing strategies for Indian farmers. In recent years, the marketing landscape has shifted increasingly toward digital marketing, enabling farmers to connect directly with consumers. This direct marketing approach not only enhances farmers' potential earnings but also increases consumer satisfaction. Therefore, understanding consumer perceptions and preferences regarding the digital marketing of agricultural produce is essential. To collect relevant data, a structured questionnaire was administered to 941 consumers in Pune city, providing insightful results. The findings show that the main benefits consumers associate with digital marketing of agricultural produce include home delivery, time savings from commuting, and the convenience of shopping based on availability. Notably, 79.2% of consumers expressed a willingness to purchase agricultural products online. They mainly expect high-quality, fresh produce to be delivered conveniently to their doorstep. To effectively promote digital marketing of agricultural produce, targeting lower-income groups through awareness campaigns and trial experiences is vital. Meanwhile, retaining higher-income consumers depends on maintaining produce quality and offering prompt, efficient online services. This research emphasizes the need for tailored marketing strategies that meet diverse consumer expectations and help Indian farmers succeed in the digital marketplace.

This study explores avocado peel (Persea americana) as a sustainable and eco-friendly biosorbent for removing Basic Blue 3 (BB3) cationic dye from water. The lignocellulosic structure of avocado peel, along with its surface carboxyl, hydroxyl, and phenolic groups, is expected to provide high adsorption capacity. Additionally, the material was subjected to temperatures from 80°C to 800°C to produce various carbonized products, and adsorption experiments were optimized using these materials. Parameters such as adsorbent amount, initial dye concentration, contact time, temperature, and agitation speed were all fine-tuned. Overall, the results suggest that avocado peel is a promising, low-cost, biodegradable alternative for dye removal from textile wastewater due to its effective adsorption capacity.

The benefits of solar power are widely known and accepted for general industrial, government and consumer use. The work in this paper demonstrates the significant benefits of the use of Photovoltaic Arrays for disaster response and compares their use to the normally deployed generators (gensets) which are commonly deployed in emergency response. In this paper, we consider the case of India, which has a large population and is highly vulnerable to natural disasters.[1] The effect of an outage caused by a disaster is serious to a society, because power distribution lines are a common basic infrastructure for life support. Everything today is dependent on power be it the basic lighting or the advanced facilities such as communication, medical services, filtered water, cooking appliances, elevators and electrical motors used to sustain daily life. Modern life is difficult without power even for few hours thus quick response in providing emergency power is one of the most crucial aspects of a successful response to disasters. In this paper, we show that portable solar energy systems can be used for prompt response to disasters thus providing aid for necessities. Renewable power also helps with fossil fuel conservation which in turn can help reduce the occurrence of natural disasters. Solar technology can be provided in a mobile format, e.g. Rapid-roll [2] which unfurls solar arrays like a carpet behind a truck which can be rapidly deployed by rolling out to generate energy in the region of catastrophe. It carries the solar panels estimated to power a mobile clinic with 120 beds, it generates an average of 11KWh of energy a day, connected to batteries that can store 24KWh for non-sunshine hours. The mobile solar power can also provide refrigeration for medical supplies, water filtration systems and communication equipment that are required for recovery process. In the case when recovery lasts a long period of time, generators will need large amount of fuel which gives solar power it’s major advantage. Clearly fossil fuel emissions are increasing at fastest rate for 7 years and even though India has the solar target of 100GW by 2022 but so far only 28GW has been installed since 2011.[3] Given that India is world’s most disaster affected country and among the top 15 countries with maximum power outage[1], it is clear that solar power is best suited for disaster relief operations to provide significant relief and to reduce the risk of future disasters realizing the solar mission target.

(This research paper is a modest effort to highlight the foreign policy of the United States towards south Asia and its nearby regions during the time of cold war. This research paper is also attempting to show the real motive of super powers in this region & what strategies they employed to achieve their goals.)

Gamification is a modern method that seeks to influence wellness and health-related frameworks positively. It is not just because gamified solutions can get individuals more engaged but can make them more responsible for their health-related assessments. This paper qualitatively analyzes the concept of the gamified solution in healthcare through systematic literature review by using PRISMA methodology to figure out the health domains and the impact on solicitation of gamification in the healthcare sector. According to WHO, non-communicable diseases (NCDs) kill 41 million people each year, equivalent to 71% of global mortality, which can be controlled through modifiable behavior. Studies proved that intrinsic motivation had got an advantage over extrinsic motivation in the case of health behaviors. Gamification is considered as a solution for motivating people both intrinsically and extrinsically. The need for more research on gamified intervention in health care still remains as a gap. The article attempts to review more about gamification as health solutions, focusing more on research articles reported in Emerald Insight and Science Direct.

The bolt -of-time marvels are everywhere in the physical world, natural frameworks furthermore, human culture. Regardless of its extraordinary significance in material science, the second law of thermodynamics can just effectively clarify little rates of these bolt- of-time wonders in the physical world. For instance, the intense second law of thermodynamics can neither clarify bolt- of-time marvels like Darwin's development which occurred in natural frameworks, nor the development which occurred in human culture. This paper sums up the second law of thermodynamics into an all-inclusive law of material science called law of harmony that is all around relevant in every framework represented by quantum mechanics inclusive of physical frameworks, organic frameworks, and human culture. Idea of entropy in factual material science is summed up utilizing, idea of relative entropy. Law of balance is one of five material science laws of sociology, which depends on another translation of quantum mechanics. This paper settle two exceptional issues in present day material science: how to sum up the second law of thermodynamics to non-balance material science, and the idea of bolt- of -time. This paper reasons that the permanent procedures and bolt- of -time marvels in the physical world, organic frameworks, and human culture are in a general sense a similar quantum wonders due to undetermined nature of quantum occasions inclusive of human decisions.

IoT (Internet of Things) is poised to transform the real world objects into intelligent virtual objects in the near future. As sensing, communication, and control become ever more complex today, this technology is applied in transforming the Internet into a fully integrated ecosystem. IoT is the next revolutionary technology in transforming the Internet into a fully integrated future ready Internet. IoT allows people and things to be connected at any time, any place, with anything and anyone, by using any path/network and any service. The advances in computer hardware, embedded system devices, networking devices, display devices, control devices, software enhancements, etc. have hugely supported IoT to grow rapidly. The main objective of this paper discusses actual implementation of IoT based monitoring of network infrastructure and its associated peripheral devices/equipment’s for parameters like temperature and power supply etc. using appropriate sensors. The main utility of the work presented, is in 24x7 monitoring of remotely located devices where it is impractical to monitor manually with direct physical presence. The implementation has been done using an integrated IoT device BOLTIoT.

This research was done with the idea to design, analyze, and compare efficiency low head turbine's hydropower energy generation capability in rural sectors of the Ganga river basin for the utilization by the local population. For the past decades, CFD ANSYS (Fluent) has been the preliminary choice of engineers for designing and analyzing the flow of fluids across structures and sections. Small Scale Hydropower (SSHP) has proven to create energy from Run of the River segment of water streams, making the world shift towards a cleaner and greener environment with minimal impacts on the aquatic ecosystem. To locate these zones of Hydropower Potential is a relatively challenging task, but with the exponential growth of Geographic Information System (GIS), the application of this system will make the challenge easier. Also with the rise in demand and to decrease the load on the central grid, a decentralized system such as solar, hydro, or wind does exist to meet with the basic needs of the local isolated areas where transportation of electricity is not feasible neither economical. The Existence of various turbines of standard design and dimensions do exist, with an efficiency of 40-70%, overheads ranging from 5 to 15 m. The study produced the conclusion that Geotagging of location having Hydro-potential can be done based on the GIS DEM data of the region having possible water heads of 0.5 meters and above in the river basin. The vortex flow design of the turbine chamber showed excellent results in the stable production of energy. Based on the data inputs on flow, Hydraulic head, and DEM, selection of turbine type was explored. The range of Head taken is 0.3 to 5 meters. Initial designing was done in AUTOCAD with simulation runs on ANSYS (FLUENT), variation parameters included intake angles, discharge variations, dimension modeling & optimization. The resulting conclusions of this study indicated that the Singular designed system with Vortex Flow and different turbine types based on geolocation specifications has 70 -80 % efficiency in the Ganga river basin with an average output power capacity of 100KW.

The search engine becomes the most important gateway for ordinary people who are looking for useful information on the web. In spite of, users might sense failure when search engines return incorrect results that do not meet their real objectives. Such inappropriateness is due in large part to the variety of users' contexts and backgrounds, as well as the uncertainty of texts. Personalized web search (PWS) is a type of search techniques which aims at providing better search results, which restricts to individual user needs. In this paper, UPS (User customizable Privacy- preserving Search) framework is proposed which generalize user profiles for each query given to user-specified privacy specification. This generalization has aims of keeping a balance between two predictive metrics that evaluate the utility of personalization and the privacy risk of exposing the user generalized profile. When considering bi-objectives problem through greedy search, it does not adequately handle the objectives dynamically or adaptively. So, to handle these goals genetic algorithm developed in this paper. The derived results will be compared based on precision and time.

The name Integral Calculus has its origin in the process of summation and the words ‘To integrate’ literally mean ‘To give the sum of.’ It was only afterward that it was seen that the subject of integration could also be viewed from the inverse of the differentiation point of view. In these classical integrals (integration with respect to measures), we must consider a measure or signed measure (i.e., additive only). This characteristic is very effective and convenient but is often too rigid. As a solution to the rigidness problem a new approach, fuzzy integrals, was proposed with the fuzzy measure (monotonic and non-monotonic), which also has the characteristics of additivity and non-additivity. In this article, I introduced a new fuzzy measure ----- SB’s 𝛍̈ -Fuzzy measure and a new fuzzy integral - SB’s fuzzy integral.

Basic Blue 3 (BB3), one of the dyes frequently used in the textile industry, is a toxic material that imparts a blue colour to aqueous environments when discharged into the natural environment without undergoing oxidation. Consequently, its discharge into the ecosystem without disposal reduces the oxygen level necessary for life by reducing light transmittance in the aqueous environment due to its color. Therefore, it negatively affects the ecosystem. In this thesis, based on this problem, the removal of toxic Basic Blue 3 (BB3) dyestuff from aqueous media was carried out by adsorption method using activated carbon and graphene adsorbents. The adsorption method was preferred due to its advantages. These advantages include low cost, rapid processing, and high removal rates. In the adsorption experiments, the optimum adsorbent dose, the starting amount of dye, the length of time they are in contact, the temperature, and the stirring speed were calculated for both adsorbents. For activated carbon; 20 mg/L, 0,05g, 60 min, 40 oC, 150 rpm, for graphene; 30 mg/L, 0,03g, 60 min, 25 oC, 180 rpm.

This paper presents novel three-wave-hybrid solutions for the (3+1)D generalized shallow water wave equation, derived via the combined application of the Hirota bilinear method and the test function approach. These solutions introduce an arbitrary function and a multitude of arbitrary constants, enriching the diversity of solutions and offering a more comprehensive view of the solution’s structural complexity. The unveiled nonlinear phenomena within the equation provide deeper insights into the underlying dynamics, which are crucial for understanding complex wave dynamics within both aquatic and atmospheric environments. The solutions have significant implications for the fields of hydrodynamics, geophysics, and atmospheric science, as they can be used to study the propagation and attenuation of waves in shallow water and the effects of natural phenomena such as tides and storms. Digital images accompanying the study partially illustrate the propagation patterns of the waves, enhancing the visual comprehension of the theoretical findings and providing a clearer understanding of the practical applications in predicting and managing wave behavior.

The paradigm shifts in education brought by the changes of the coronavirus disease prompted educational institutions to shift from traditional face-to-face classrooms to modular and distance learning. This move also drives institutions to utilize free learning management systems and even develop such. Thus, ESSU Salcedo, through the College of Computer Studies, developed a user-friendly learning management system (LMS) named e-SSUKwela, which serves as an online classroom where lectures, quizzes, assignments, and other assessments are done. This customized LMS provides an exchange of educational resources flexible, convenient, and engaging. This research follows a Waterfall model approach in developing the desired system, where requirement analysis, acceptance test, unit test, integration test, and coding are significant parts of the software development. The system had undergone three testing strategies: Benchmark, Alpha, and Beta Tests using the ISO/IEC 9126 metrics. Results showed that in terms of usability, the application scored 88.75, which is above average. After successful testing stages on the functionality of the developed LMS, the researchers concluded that it is ready for implementation. The results indicate consistent improvement from the various testing phases across all criteria. The steady progression suggests that usability, reliability, and maintainability were key focus areas during software development. As such, it is recommended that this LMS be part of the smart campus implementation of the university; an orientation be given to faculty and students who would utilize the system; and there should be a study on the effectiveness of the system in the blended learning environment.

One of the most fascinating aspects that Indian classical music is, it connects the musical part with human emotions through its flawless structure. The music incorporates various tonal patterns, improvisations, and intricate microtonal subdivisions. The backbone of this musical system is totally dependent on ragas. Determining these ragas is like finding a needle in a haystack, which is not amenable to conventional genre classification methods. Although the domains of Machine Learning and Deep Learning have significantly enhanced Music Information Retrieval (MIR), their application in raga identification in Indian classical music is still a pioneering area with only a few scattered works. This survey offers an exhaustive view of the recent Machine Learning methods for raga recognition. It talks about the traditional feature extraction methods like Mel Frequency Cepstral Coefficients (MFCC), Chroma features, and Spectral Centroid, and advanced Deep Neural Network architectures such as Convolutional Neural Networks (CNNs), ResNet, and cutting-edge models like Wav2Vec2.0 and the Audio Spectrogram Transformer (AST). The paper highlights significant research gaps in the standardization of datasets and the understanding of methods for modeling, interpreting, and generalizing musical instruments. It looks back to the evolution of computational raga modeling while acknowledging the challenges that have persisted and suggesting some future directions based on a mixture of symbolic, acoustic, and affective elements. The study presented, by aligning artificial intelligence with India's classical music heritage, intends to build data-driven systems that are reverent of tradition and progressive in technology.

The Internet of Things (IoT) has seen rapid growth, leading to a surge in connected devices that often come with limited computing power, memory, and energy resources. Ensuring secure communication and reliable authentication in these resource-constrained environments remains a significant challenge. This paper outlines the design and performance evaluation of a lightweight mutual authentication protocol tailored specifically for IoT devices with constraints. The proposed protocol employs cost-effective cryptographic techniques and requires only a few message exchanges to establish mutual authentication between two communicating parties, all while ensuring confidentiality, integrity, and protection against common security threats like replay attacks, impersonation, and man-in-the-middle attacks. A comprehensive system model and set of assumptions are established to accurately reflect the operational context of IoT networks. The protocol design and discussion delve into each step of the authentication process, highlighting its speed and security. A detailed performance analysis, supported by analytical modeling and comparisons with other lightweight authentication schemes, demonstrates that the proposed protocol reduces execution time, communication overhead, and energy consumption without compromising security. The results indicate that this protocol is highly suitable for large-scale IoT implementations, particularly in scenarios that require high efficiency and low power usage. The paper concludes by outlining future research directions, including enhancing system scalability, compatibility with various IoT frameworks, and integrating lightweight cryptographic mechanisms that are resilient against quantum computing threats.

The article examines and analyzes the use of cosmonyms within the language of the people. It identifies how cosmonyms are represented in Nakhchivan folklore texts and clarifies their functional role in the cultural consciousness of the population. During the research, references were also made to the Holy Qur’an. It was revealed that several of these cosmonyms appear in various surahs of the Qur’an. Therefore, the article discusses these instances in detail and provides illustrative examples from the holy text. The presence of cosmonyms in Qur’anic verses contributes to their sacred status among the people. Just as the Sun, Moon, and stars adorn the sky, folklore genres enrich our national cultural heritage. Certain verses in the Qur’an that mention cosmonyms find parallels in folklore narratives, and the article demonstrates this through selected examples. Additionally, folklore texts include personal names derived from cosmonyms – such as Sun, Star, Moon, Crescent, and others – names that are still actively used today and cannot be considered archaic. Anthroponyms based on these cosmonyms are also analyzed within the article. The study further explores the beliefs and traditional views of our ancestors related to cosmonyms, summarizing and classifying them. It is noteworthy that concepts associated with the Sun, Moon, and stars remain meaningful in contemporary society, and people continue to place significant value on these beliefs. The article supports this claim with multiple examples.

Part-of-speech (POS) tagging is a fundamental preliminary task in Natural Language Processing (NLP), which assigns grammatical tags to words in a sentence, such as nouns, verbs, and adjectives. In lowresourced languages like Gujarati, few annotated datasets and few linguistic tools hinder the establishment of accurate part-of-speech tagging systems. This paper introduces a part-of-speech tagging framework for the Gujarati language utilizing a Transformer-based IndicBERT model. The Precision and recall ensure accurate sequence dependencies and tag-to-tag consistency, and the model is enhanced with a Conditional Random Field (CRF) layer. The model was trained and validated on the Universal Dependencies Gujarati Treebank (GujTB) dataset containing approximately 6200 sentences or 125,000 tokens. Evaluation indicates accuracy of 97.12% precision of 96.90% , recall of 96.60%, and F1-score of 96.75%, outperforming classical assessment methods like Hidden Markov Model (HMM), CRF, and BiLSTM baselines. The introduced contexts and contextual embedding layer allow for a better representation of long-range dependencies and language variations thus making the model highly applicable for Gujarati part-of-speech tagging and potentially can be utilized as a mechanism for other Indic languages tasks.

Transfer learning has become a crucial technique in machine learning, allowing models that have been trained in one domain to be effectively utilized for new, related tasks with limited labeled data. Leveraging insights from pre-trained models, transfer learning enhances model generalization, reduces training times, and mitigates challenges associated with limited data availability. This approach has been widely adopted in fields like computer vision, natural language processing, and healthcare, where the collection of extensive, annotated datasets can be particularly difficult. A notable benefit of transfer learning is its capacity to leverage deep learning models that have been pre-trained on extensive datasets, like ImageNet for visual tasks or GPT for text analysis, for specialized applications. This has led to significant advancements in fields such as medical diagnosis, speech recognition, and autonomous systems. Nevertheless, cross-domain adaptation also faces several hurdles. One major challenge is domain shift, where discrepancies in data distributions between the source and target domains can impede performance. Furthermore, negative transfer may arise when the knowledge from the source domain does not align well with the target domain, resulting in less effective learning. Ensuring proper feature alignment across domains is another challenge, as differences in data characteristics can lead to inconsistencies in model outputs. Researchers have developed advanced methodologies to address these challenges, including domain adaptation, adversarial learning, contrastive learning, and self-supervised techniques. These methods enhance model robustness and adaptability, making transfer learning more effective in practical applications. However, the quest for scalable, interpretable, and unbiased transfer learning models continues to be a significant area of research. This study explores the key applications, techniques, and obstacles related to transfer learning, offering perspectives on how advancements can enhance cross-domain adaptation, leading to more efficient, flexible, and generalizable AI solutions.

With the exponential growth of the internet, cyber threats such as malicious URLs have emerged as a critical challenge in cybersecurity. Traditional blacklisting methods fail to detect newly generated or obfuscated URLs. This paper introduces an advanced hybrid deep learning and ensemble-based framework for detecting malicious URLs. The proposed system combines Convolutional Neural Networks (CNN) for lexical feature learning with a Random Forest ensemble to refine classification accuracy. The hybrid model captures structural, lexical, and behavioral attributes of URLs, enabling a comprehensive analysis. The proposed approach achieves superior accuracy, precision, and recall compared to conventional methods. The system is deployed via a Flask-based web application that performs real-time analysis and classification of URLs as safe or malicious. Experimental results demonstrate a detection accuracy of 98.7%, showcasing the potential of hybrid AI models in fortifying modern cybersecurity frameworks.

Heart disease is one of the leading causes of mortality worldwide, and early diagnosis can significantly improve patient outcomes. Traditional methods of diagnosis rely on manual analysis of medical parameters, which can be time-consuming and prone to errors. This project aims to develop a machine learningbased predictive model that can assist in identifying individuals at risk of heart disease based on clinical and lifestyle data. The model is trained using datasets containing patient information, including age, blood pressure, cholesterol levels, heart rate, and other medical parameters. Various machine learning algorithms are employed to analyse the data and predict the likelihood of heart disease. The performance of these models is evaluated using metrics like accuracy, precision, recall, and the ROC-AUC score. The developed system can be deployed as a web or mobile application to provide a user-friendly interface for real-time predictions. The goal of this project is to assist healthcare professionals and individuals in making early, data-driven decisions regarding heart health. Future enhancements may include the integration of deep learning techniques and real-time IoTbased health monitoring for improved accuracy and usability.