Volume 14 | Issue 4 |

 Abstract

Generative Artificial Intelligence has advanced rapidly because of transformer based Large Language Models (LLMs), which can generate fluent text, answer questions, summarize content, and support many language tasks. At the same time, these models still face important limitations. They may produce hallucinated answers, rely on static training data, and fail to access up to date or domain specific information. Retrieval Augmented Generation (RAG) offers a practical way to address these issues by retrieving relevant external evidence before generating a response. This paper presents a structured study of the development of generative AI, the working principle of LLMs, and the role of RAG in improving factual reliability. It also explains the supporting technologies used in RAG systems, such as semantic embeddings, vector search, dense passage retrieval, and sparse retrieval methods. A comparative discussion is included to show how traditional QA systems, standalone LLMs, and RAG based systems differ in flexibility, accuracy, and trustworthiness. The study concludes that RAG is one of the most practical directions for building more reliable generative AI systems, especially in settings where current and verifiable knowledge matters .

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Generative Artificial Intelligence, Large Language Models, Transformer Architecture, Retrieval-Augmented Generation, Natural Language Processing, Embeddings, Vector Databases, Dense Retrieval, Hallucination Reduction, Knowledge Grounding, AI Systems, Information Retrieval.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Reading is a fundamental skill that influences academic achievement across all subjects. However, many middle school learners struggle with reading due to limitations in phonics, vocabulary, fluency, and comprehension. This paper reviews existing literature on reading difficulties among middle school students, with a focus on multilingual contexts such as India. The review identifies major challenges faced by learners and highlights gaps in existing research. The findings indicate a need for structured reading interventions at the middle school level.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Reading Difficulties, Fluency, Vocabulary, Comprehension, Middle School Learners

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Artificial Intelligence (AI) has increasingly entered the field of mental health care and counseling psychology. Digital tools such as conversational agents, AI-driven mental health platforms, and automated therapeutic applications are now being used to provide psychological support, psychoeducation, and early intervention. While these technologies promise accessibility, affordability, and scalability, there are growing concerns about whether AI can preserve the deeply personalized nature of psychological counseling. Personalized intervention in counseling relies on empathy, emotional attunement, therapeutic alliance, and the counselor's ability to adapt techniques to the client's unique psychological and cultural context. AI systems, however, operate primarily through algorithmic decision making and programmed responses. Although they may successfully deliver structured techniques such as cognitive behavioral therapy, they may lack the nuanced emotional sensitivity that human counselors bring to therapy. The present study examined whether the integration of AI in counseling psychology may reduce the level of personalized intervention in therapy. The study was conducted over a period of 45 days and focused on young psychologists between the ages of 25 and 45 years, including both male and female participants. A quantitative research design was used to explore psychologists' perceptions regarding AI-assisted counseling, empathy in therapy, therapeutic alliance, and personalization of intervention. The theoretical framework for this research was grounded in Carl Rogers' Person-Centered Therapy, Bordin's Therapeutic Alliance Model, and the Technology Acceptance Model. Findings suggest that although AI tools may improve accessibility and provide supplementary support in mental health services, psychologists express concerns regarding the potential loss of empathy and individualized therapeutic processes when AI is heavily relied upon. The study highlights the importance of integrating AI in a balanced manner where technology supports but does not replace the humanistic foundation of counseling psychology.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Artificial Intelligence, Counseling Psychology, Therapeutic Alliance, Personalization, Mental Health

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study investigates the effectiveness of a structured reading intervention programme in improving English reading skills among Class VII students. A quasi-experimental design using pre-test and post-test was adopted. The intervention included phonics instruction, vocabulary activities, guided reading, and repeated reading. The results showed significant improvement in students' reading performance. The study highlights the importance of structured teaching strategies in enhancing reading skills.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Reading Intervention, Fluency, Comprehension, Vocabulary, Middle School

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The present study aims to develop a functional dairy product--herbal curd--fortified with antidiabetic medicinal plants and to evaluate its probiotic properties, phytochemical composition, nutritional profile and sensory acceptability. Herbs used in the study were Glycyrrhiza glabra (Mulethi), Gymnema sylvestre (Gudmar), Stevia rebaudiana (Madhuparni), Beta vulgaris (Beetroot), cinnamon, cardamom, dry ginger and nutmeg. Curds were prepared by supplementing milk with herb powders followed by fermentation using a curd starter culture. The prepared curd was evaluated for sensory properties, phytochemical composition, nutritional value and probiotic characteristics. Lactic acid bacteria (LAB) were isolated and characterized using morphological methods. The isolates exhibited tolerance to low pH (2-3), bile salts and showed antimicrobial activity against Staphylococcus aureus and Escherichia coli. Phytochemical analysis confirmed the presence of alkaloids, flavonoids, phenols and tannins. Nutritional analysis revealed compliance with FSSAI standards. The study demonstrates that herbal curd can serve as a potential functional food with antidiabetic and probiotic benefits.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Herbal curd, Probiotics, Antidiabetic activity, Lactic acid bacteria, Phytochemicals, Functional food _______________________________________________________________________________________________

  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

Person Re-Identification, Market-1501, Browser Based Vision, Canvas API, Pixel Similarity, Surveillance, Computer Vision, Deep Learning

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The financial markets have been transformed from a traditional market ecosystem to digital market systems. Traditional market systems were characterised by manual trading, delayed information dissemination, and higher transaction costs. The evolution of digital finance transformed the ecosystem through electronic real-time trading, smartphone trading apps, and speedy information transfer. Digital finance paves the way for increased retail participation. The digital finance ecosystem enables transaction transparency through depositories, thereby reducing settlement cycles and speeding up information processing. Since information is an integral part of the financial market, the persistence of information asymmetry creates an imbalance, as some investors who possess it can earn abnormal returns. These imbalances can lead to potential mispricing in financial markets. Weak-form market efficiency asserts that there is no information asymmetry regarding past prices, since they are publicly available and already reflected in current security prices. The paper examines how digital finance systems influence weak-form market efficiency by speeding up information dissemination and eliminating abnormal gains driven by predictability.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

digital finance, information asymmetry, information dissemination, weak-form market efficiency.

  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

Keywords: Physical Restraints, Structured Teaching Programme, Knowledge, Attitude, Staff Nurses

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Deep learning has emerged as a transformative approach within Artificial Intelligence and Machine Learning, significantly enhancing the capabilities of image and pattern recognition systems in data science. This work explores advanced deep learning architectures such as Convolutional Neural Network, Recurrent Neural Network, and Transformer architecture, focusing on their ability to automatically learn complex and hierarchical representations from large-scale datasets.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Deep Learning Approaches for Image and Pattern Recognition in Data Science

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The Junglemahal region of southwestern West Bengal, particularly Jhargram district, experiences recurrent forest fires with significant ecological and socioeconomic consequences. This research note synthesizes available evidence from government reports, forest department communications, and recent fire alert data to characterize the patterns, causes, and impacts of forest fires in this vulnerable landscape. Findings indicate that Jhargram consistently ranks among the most fire-prone districts in West Bengal, with fires being predominantly human-induced rather than natural phenomena. Local communities deliberately set ground vegetation ablaze for multiple purposes, including collection of Mohul flowers (Madhuca longifolia), facilitating minor forest produce gathering, clearing undergrowth for easy movement, and occasionally for poaching small wildlife. The study highlights a paradox: while total fire incidents in West Bengal declined by approximately 60% from 2022-23 to 2023-24, Jhargram remains a persistent hotspot requiring targeted interventions. This research advocates for community-based fire management approaches that address the underlying socioeconomic drivers rather than relying solely on suppression-oriented strategies.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Forest fires, Junglemahal, Jhargram, anthropogenic fires, fire-prone areas, West Bengal

  License

Creative Commons Attribution 4.0 and The Open Definition