345 10 3494 https://omeka.ibu.edu.ba/files/original/92ae89ad7a41ba9c5ec9a5571ee571f8.pdf bf338fbd29dfcc0e55ec2cf5e3c1ed98 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource MODEL FOR PREDICTION OF LUNG CANCER Author Author Lamija Šetić Abstract A summary of the resource. Lung cancer remains one of the leading causes of cancer-related deaths globally, with early detection being critical to increasing survival rates. The primary goal of this project is to design and implement a machine learning classification model capable of accurately identifying lung cancer based on patient data. This work utilizes a publicly available dataset which includes features such as age, gender, air pollution levels, smoking habits, and other relevant health indicators. The methodology involves preprocessing the dataset to handle missing values, normalize input features, and encode categorical variables. Various classification algorithms were explored, including Logistic Regression, Random Forest, Support Vector Machine (SVM), and Gradient Boosting, to determine the most effective model. Model performance was evaluated using standard metrics such as accuracy, precision, recall, and F1-score through cross-validation techniques to ensure robustness. Initial results indicate that ensemble methods, particularly Random Forest and Gradient Boosting, significantly outperform other models, achieving an accuracy of over 96%. These findings suggest that machine learning techniques can play a crucial role in assisting medical professionals with early diagnosis, thereby contributing to timely treatment and improved patient outcomes. In conclusion, this project demonstrates the effectiveness of supervised machine learning algorithms in medical data analysis and highlights the potential of data-driven solutions for real-world health challenges. Future improvements may involve integrating additional medical features and deploying the model in a web-based diagnostic tool for practical use. Keywords Keywords. senior design project, lung cancer classification, machine learning, Random Forest, early diagnosis https://omeka.ibu.edu.ba/files/original/38241da3fc5ef5b9474be1c93fe89fa7.pdf 6bba05f2040411ac0ec700b0bff6b369 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Quantitative Analysis of Voice Recognition Models Author Author Faris Muhović Abstract A summary of the resource. With the growing adoption of virtual communication and voice-driven applications, the need for accurate, real-time, and privacy-conscious transcription tools has become critical. Existing solutions largely rely on cloud infrastructure, introducing concerns around latency, cost, and data privacy. This project investigates whether modern speech recognition models can perform competitively in fully offline environments while maintaining accuracy and responsiveness. To this end, we conducted a comparative evaluation of four voice transcription model, Whisper, Faster-Whisper, Wav2Vec2, and Vosk, using the AMI Meeting Corpus. Each model was assessed based on four key metrics: Word Error Rate (WER), Character Error Rate (CER), BLEU, and ROUGE-L. Our findings demonstrate that Faster-Whisper outperforms the others in accuracy and latency, making it a strong candidate for edge deployment. Building upon this analysis, a lightweight desktop application was developed using Python and PyQt5. The app captures microphone input in real time, applies VAD (Voice Activity Detection) and loudness filtering to reduce noise, and transcribes valid segments using Faster-Whisper. Additionally, the tool integrates Ollam, a local LLM engine to optionally generate intelligent responses to transcribed text. This work contributes a dual outcome: a detailed empirical evaluation of modern transcription models on realistic meeting audio, and a functional, privacy-preserving voice assistant prototype for local systems. The results highlight the feasibility and value of running sophisticated voice AI tools on personal machines without cloud dependency, paving the way for secure adoption in sensitive domains such as legal, healthcare, and enterprise communication. Keywords Keywords. speech recognition, Whisper, Faster-Whisper, transcription models, real-time, privacy, WER, PyQt5 machine learning sdp https://omeka.ibu.edu.ba/files/original/9c29e4d40232e01702af619983abfb59.pdf dc20133763eb478f485ac342c6add442 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Movie Recommender Web Application Author Author Salih Rogo Abstract A summary of the resource. Problem Statement: The exponential growth of digital content has created an information overload problem, making it increasingly difficult for users to discover relevant movies from vast catalogs. Traditional browsing methods are inefficient and fail to leverage user preferences and behavioral patterns, necessitating intelligent recommendation systems that can provide personalized movie suggestions. Methods and Procedures: This project developed a comprehensive movie recommendation system utilizing collaborative filtering techniques, implemented with a Python FastAPI backend and a Remix.js frontend. The system employs the Singular Value Decomposition (SVD) algorithm, trained on the MovieLens 32M dataset, which contains 162,541 users and 59,047 movies. The architecture integrates multiple data sources, including IMDb metadata through the OMDb API, implements RESTful API endpoints for recommendation generation, and provides a modern web interface for user interaction. The system was deployed on Heroku with MySQL database hosting on the Railway platform. You can visit the recommender system by yourself on the following URL: www.salihrogo.me Results: Comprehensive evaluation demonstrated solid performance across key metrics: Mean Absolute Error of 0.82, indicating good predictive accuracy, Hit Rate of 58.7% showing effective recommendation relevance, and catalog coverage of 72.3% ensuring adequate movie variety. The system achieved 86.4% user coverage, minimizing cold start problems, while maintaining a diversity score of 0.612 and a novelty score of 0.578, indicating balanced recommendations between popular and lesser-known content. Testing suite comprising 43 test cases validated system reliability across unit, integration, and end-to-end scenarios. Conclusion: The implemented movie recommender system successfully addresses the content discovery challenge through effective collaborative filtering, demonstrating production-ready performance with clear pathways for future enhancement. The system provides a scalable foundation for personalized movie recommendations while maintaining data integrity, security, and user experience standards. Keywords Keywords. collaborative filtering, movie recommendation system, SVD algorithm, MovieLens dataset, FastAPI, machine learning, personalized recommendations, web application sdp web development https://omeka.ibu.edu.ba/files/original/7376fc04d82c9d988bdd80981021285a.pdf 8749db0a6d643a6aa4fac76ddda611bc Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource THE ROLE OF NLP IN DETECTING HATE SPEECH ON SOCIAL MEDIA Author Author Nora Žehak Abstract A summary of the resource. The accumulation of user-generated content on social media has increased the presence of hate speech and offensive language online, causing serious psychological impact on people. This presents a critical problem for platform moderators and software developers that are trying to create safer online communities. The main purpose of this research is to detect and classify hate speech in social media posts using Natural Language Processing and Machine Learning techniques. By improving the reliability and efficiency of harmful content detection, this research focuses to support efforts in content moderation and reduce the influence of toxic online interactions. The problem was faced using a multi-class classification approach to categorize text into hate speech, offensive language, or neutral speech. To achieve this, a publicly available dataset of labeled tweets was processed through a detailed pipeline involving data cleaning, normalization, and transformation. Text preprocessing steps such as removing URLs, mentions, hashtags, and punctuation, followed by tokenization, stopword removal, and lemmatization, were essential to reduce noise and standardize the input. The cleaned data was then vectorized using Term Frequency-Inverse Document Frequency (TF-IDF) to represent tweets numerically, enabling machine learning algorithms to extract meaningful patterns. The preprocessed dataset was used to create and train a number of classification models, including Logistic Regression, Support Vector Machines, Random Forest, Naive Bayes, and many boosting methods like XGBoost, LightGBM, and CatBoost. The Synthetic Minority Over-sampling Technique (SMOTE) was used during training to address class imbalance and improve the model's detection of minority classes, including hate speech. A broad range of metrics, such as accuracy, precision, recall, F1-score, and confusion matrices, were used to assess performance in order to provide an objective and fair assessment for each class. By demonstrating how NLP and ML techniques may be used to more efficiently detect hate speech, this research supports continuing attempts to automate content moderation. It also lays the foundation for future developments in the discipline by identifying important difficulties in managing complex language and contextual uncertainty. Keywords Keywords. hate speech detection, natural language processing, machine learning, text classification, content moderation, offensive language, TF-IDF, supervised learning, data preprocessing, XGBoost, Random Forest, LightGBM, SVM, Logistic Regression https://omeka.ibu.edu.ba/files/original/c99f4e38c2dcc97c5a3b0ca83c326795.pdf 734d69b81ac5f0a793d17702b63a259c Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Retrieval-Augmented Generation System for Efficient Access to University Information Author Author Aldin Ahmethodžić Abstract A summary of the resource. Accessing university-related information, such as course syllabi, class schedules, and announcements, can be inefficient and fragmented, especially when data is distributed across multiple platforms. This often leads to time-consuming searches and a suboptimal user experience for both students and staff. This project presents a solution in the form of an intelligent, centralized platform capable of understanding natural language queries and delivering accurate, contextually relevant responses. It implements a Retrieval-Augmented Generation (RAG) system designed to streamline information access at International Burch University. The system features a multi-layered architecture combining semantic search and generative AI. Vespa.ai serves as the vector database enabling high-performance similarity search, while OpenAI’s GPT models handle natural language understanding and generation. MongoDB is used for session tracking and user state management. Documents in PDF, TXT, and web-based formats are ingested through a pipeline that performs scraping, text extraction, and chunking using LangChain’s RecursiveCharacterTextSplitter. The system is accessible via a modern web interface built with React.js, Vite, and Tailwind CSS, and also exposes a FastAPI-based REST API for backend interaction. Secure access is enforced using JWT-based authentication and authorization. The chatbot component retrieves relevant context through semantic search and maintains coherent multi-turn conversations using conversation history tracking. The developed system significantly improves the accessibility and efficiency of retrieving university information. It provides fast, context-aware responses to user queries by combining robust retrieval techniques with generative language models. The web-based interface ensures ease of use for both technical and non-technical users, while the modular backend supports scalability and maintainability. Intelligent query rephrasing and optimized chunk retrieval contribute to improved precision and user satisfaction. Overall, this solution demonstrates how RAG-based systems can transform information access in academic environments by offering a centralized, intelligent platform tailored to users' needs. Keywords Keywords. Retrieval-Augmented Generation, Semantic Search, Vespa.ai, OpenAI GPT, FastAPI, React.js, MongoDB, University Information System, Natural Language Processing natural language processing sdp https://omeka.ibu.edu.ba/files/original/b405bccb488e75bf8fb55f27f34023be.pdf a77087e12a6314709f1c2dad1d73a655 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Master's Theses Description An account of the resource IT Master's Thesis collection features master's theses authored by graduate students in the Department of Information Technology. Each thesis reflects a significant research effort, combining theoretical knowledge with practical application to address complex challenges in the IT domain. These works demonstrate students’ advanced understanding of information systems, software engineering, data science, cybersecurity, and emerging technologies. The theses serve as a testament to the students' capability to conduct independent research, propose innovative solutions, and contribute to the advancement of the IT field. Subject The topic of the resource IT Master's Thesis Language A language of the resource English Language Author Author IT Department Master’s Students Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Sentiment Analysis Techniques and Applications in the News Articles Author Author Mirza Novalić Abstract A summary of the resource. Sentiment analysis is essential for understanding public opinion, especially in the context of news articles, where tone and sentiment can significantly impact and control readers' perception and understanding of the content. This study explores a variety of sentiment analysis techniques that are applied to a vast amount of articles gathered from “New York Times” in the past two decades. The research focuses on the performance of traditional machine learning models, deep learning models and hybrid approaches. The aim of the paper is to answer three key questions regarding which approach is the most suitable for this problem and how fine-tuning affects end results.<br /><br />To address these questions, throughout the research, traditional machine learning models including Naive Bayes, Linear Support Vector Classification (SVC) and Logistic Regression were implemented. Among these approaches, Linear SVC achieved the best scores across all evaluation metrics. In the deep learning category, Long Short-Term Memory (LSTM) networks were applied. This approach provided exceptional performance which was overall better than traditional models. RNNs scored similarly as Linear SVC, while outperforming other traditional algorithms. <br /><br />A hybrid approach including the BERT model was another method that was explored, which combined specific architecture with deep learning-based contextual understanding. The results demonstrated high classification results, which supports the hypothesis that hybrid models can increase performance of sentiment prediction. Furthermore, fine-tuning of different models improved their performance, which highlights the importance of optimizing pretrained models for specific types of analysis. <br /><br />Overall, the findings confirm that deep learning models usually outperform traditional variants of machine learning methods while hybrid models can offer additional potential and perspective for enhancing sentiment classification in news articles. The study provides deep and valuable insights into effectiveness of different sentiment analysis and natural language processing (NLP) techniques, while at the same time discussing new possibilities and improvements in the field. Keywords Keywords. Sentiment Analysis, Traditional Algorithms, Deep Learning, Fine-Tuning, Hybrid Approach deep learning fine-tuning hybrid approach sentiment analysis Traditional Algorithms https://omeka.ibu.edu.ba/files/original/85f02200d8b6a059ec3861f417da6e72.pdf 1f82ffbbe02c17154c3399fe61ce48b1 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Master's Theses Description An account of the resource IT Master's Thesis collection features master's theses authored by graduate students in the Department of Information Technology. Each thesis reflects a significant research effort, combining theoretical knowledge with practical application to address complex challenges in the IT domain. These works demonstrate students’ advanced understanding of information systems, software engineering, data science, cybersecurity, and emerging technologies. The theses serve as a testament to the students' capability to conduct independent research, propose innovative solutions, and contribute to the advancement of the IT field. Subject The topic of the resource IT Master's Thesis Language A language of the resource English Language Author Author IT Department Master’s Students Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Sentiment Analysis And Price Prediction For Accommodation Reviews in Bosnia And Herzegovina: A Comparative Study of NLTK and Hugging Face NLP Techniques Author Author Amila Čaušević Abstract A summary of the resource. The growing field of natural language processing (NLP) has huge potential in the advancement of consumer feedback and its application in determining pricing strategy in the hospitality industry. In this thesis, sentiment analysis and price predictions of accommodation reviews in Bosnia and Herzegovina are analyzed through a comparative study of two of the most commonly used approaches in NLP: NLTK - representing traditional methods, and Hugging Face - representing modern techniques. Initially, a long process of text preprocessing is performed that includes tokenization, lemmatization, stopword removal, and filtering of positive and negative reviews. Quantitative analysis such as word frequency distributions, measures of lexical diversity, and word co-occurrence tests reveal patterns within language use as well as the relationship between review attributes and sentiment.<br /><br />Different frameworks  for sentiment analysis are then compared. The Hugging Face sentiment pipeline and more modern and recent transformer architectures like BERT, RoBERTa, and XLNet are compared with more traditional techniques (e.g., NLTK/VADER). Metrics for evaluation such as accuracy, precision, recall, and F1-score are used to assess the performance of the sentiment models. In order to develop predictive price models based on regression techniques like Linear Regression, Random Forest, and Gradient Boosting, the thesis additionally integrates sentiment scores with quantitative metadata, such as review ratings, location ratings, and accommodation categories. The results show that Random Forest regression is the most effective method for identifying subtle, non-linear sentiment-price correlations, even though transformer-based sentiment analysis can show promise in identifying subtle signals within guest reviews. Last but not least, this work offers helpful recommendations to help hoteliers in Bosnia and Herzegovina to create focused pricing strategies while also enhancing the general guest experience. Keywords Keywords. NLP, Sentiment Analysis, Accommodation Reviews, Price Prediction, Bosnia and Herzegovina Accommodation Reviews Bosnia and Herzegovina NLP Price Prediction sentiment analysis https://omeka.ibu.edu.ba/files/original/1e77e2888b1af34ce1b04e8c6c8b1578.pdf 3102207937ea6ff2b0c16d894cdecc49 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Master's Theses Description An account of the resource IT Master's Thesis collection features master's theses authored by graduate students in the Department of Information Technology. Each thesis reflects a significant research effort, combining theoretical knowledge with practical application to address complex challenges in the IT domain. These works demonstrate students’ advanced understanding of information systems, software engineering, data science, cybersecurity, and emerging technologies. The theses serve as a testament to the students' capability to conduct independent research, propose innovative solutions, and contribute to the advancement of the IT field. Subject The topic of the resource IT Master's Thesis Language A language of the resource English Language Author Author IT Department Master’s Students Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Comparative Analysis of Machine Learning Algorithms for Real Estate Price Prediction: Bosnia and Herzegovina vs. USA Author Author Adnan Krndžija Abstract A summary of the resource. Real estate markets are impacted by a variety of variables, including changes in the population, urban development projects, and changes in economic policy. This thesis sets out to investigate the effectiveness of machine learning algorithms in predicting real estate prices, paying close attention to the particular circumstances of Bosnia and Herzegovina as well as the United States. While the US real estate market has a long history and is well-known for its capacity to bounce back from downturns in the economy, the tale of the BiH real estate industry is very different. In contrast to the United States, which has seen centuries of economic expansion, financial crises, and legislative changes, Bosnia and Herzegovina's market development is a result of a combination of past influences and present difficulties. Beyond simple quantitative comparisons, our research takes a holistic method to uncover the predictive capability of machine learning models.<br /><br />We explore the complexities of random forests and decision trees, making use of their ability to reveal intricate patterns in real estate databases. This research also includes time series modeling to recognize and comprehend the evolving patterns that characterize real estate dynamics throughout time. The analysis of SARIMAX, ARIMA, and Holt-Winters time-series models shows ARIMA's consistent accuracy, while SARIMAX and Holt-Winters excel in stability and trend capture, respectively. In machine learning, Decision Trees offer interpretability, while Random Forests show reduced error rates and enhanced accuracy. In the US dataset, SARIMAX has a Mean Absolute Percentage Error (MAPE) of 3.35% and ARIMA achieves 1.66%, while Holt-Winters shows 3.54%. Decision Trees have a MAPE of 2.97%, and Random Forests achieve 2.10%. In the BiH dataset, SARIMAX has a MAPE of 5.08%, ARIMA achieves 1.22%, while Holt-Winters shows 2.17%. Decision Trees have a MAPE of 0.83%, and Random Forests achieve 0.82%. Keywords Keywords. Real Estate, Prediction, Bosnia and Herzegovina, United States Bosnia and Herzegovina master's thesis prediction Real Estate United States https://omeka.ibu.edu.ba/files/original/898d3390481715985e1899366de2077e.pdf c83f8b36a7dcda7b6bb84552da258b06 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Master's Theses Description An account of the resource IT Master's Thesis collection features master's theses authored by graduate students in the Department of Information Technology. Each thesis reflects a significant research effort, combining theoretical knowledge with practical application to address complex challenges in the IT domain. These works demonstrate students’ advanced understanding of information systems, software engineering, data science, cybersecurity, and emerging technologies. The theses serve as a testament to the students' capability to conduct independent research, propose innovative solutions, and contribute to the advancement of the IT field. Subject The topic of the resource IT Master's Thesis Language A language of the resource English Language Author Author IT Department Master’s Students Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Optimization of Email Marketing Campaigns Leveraging Machine Learning Techniques Author Author Nejla Zukorlić Abstract A summary of the resource. Email marketing is widely recognized as an effective digital marketing channel, offering a considerable return on investment (ROI). One key challenge is determining the optimal day and time to send emails to maximize customer response rates. This thesis explores the application of machine learning (ML) algorithms to predict the best send times for email marketing campaigns, focusing on improving response rates. The research utilizes historical email marketing data, including customer demographics, response behavior, and email send dates. Based on this data, various machine learning models, including decision trees and random forests, as well as ensemble methods at the end, will be used to predict the optimal day for sending emails. The study will also examine how factors like customer age and tenure influence response rates at different times. The question is if the machine learning-based predictions of the optimal send day and time will significantly improve response rates compared to traditional methods. Also, incorporating demographic factors, such as age and tenure, hopefully will improve the accuracy of these predictions. The expected outcome is that MLbased optimization will outperform traditional scheduling methods, providing a more effective and data-driven strategy for email campaign timing. Keywords Keywords. Email Marketing, Machine Learning, Send-Time Optimization, Customer Engagement, Response Rates, Data-Driven Marketing Customer Engagement Data-Driven Marketing Email Marketing machine learning master's thesis Response Rates Send-Time Optimization https://omeka.ibu.edu.ba/files/original/7f17ded94652315afc7fbd922bcea059.pdf 98a134d0ae38d8a4fefd574989346b94 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource IT Senior Design Projects Description An account of the resource The IT Senior Design Projects (SDPs) category showcases innovative and practical final-year capstone projects developed by undergraduate and graduate students in the field of Information Technology. These projects represent the culmination of students' academic and technical expertise, demonstrating their ability to solve real-world problems through software and hardware solutions. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource AI STORYBOOK GENERATOR Author Author Ajla Čakić Abstract A summary of the resource. The rapid advancement of artificial intelligence has opened new possibilities for educational and creative applications. This project presents the design and development of an AI-powered Storybook Generator—a full-stack web application that allows users to generate personalized children’s storybooks using natural language prompts. The goal is to make storytelling more interactive, accessible, and creatively empowering by automating story creation, illustration, and narration through AI. The application addresses the problem of limited access to personalized, diverse, and engaging story content, especially for children from various cultural and linguistic backgrounds. By integrating advanced AI services, the platform generates unique stories based on user-provided inputs such as title, age group, genre, and illustration style. Using technologies like Google AI, Hugging Face’s text generation models, and custom text-to-speech tools, the system delivers cohesive narratives paired with AI-generated visuals and narration. The backend is developed using Node.js and Express.js, while the frontend is built with React.js, offering a responsive and user-friendly interface. Data is securely stored and managed using a Drizzle ORM with a PostgreSQL database. Results show that the system can produce high-quality storybooks with consistent plots, age-appropriate language, and stylized imagery, enhancing the reading experience. The platform also includes features like story exploration, user account management, coin-based generation limits, and payment integration for purchasing additional credits. In conclusion, the AI Storybook Generator showcases the potential of merging generative AI with interactive design to promote literacy and creativity in young users. Future work may involve mobile app development, multilingual support, story sharing features, and integration of educational objectives into story structure. The application serves as a scalable, customizable tool for families, educators, and storytellers worldwide. Keywords Keywords. AI story generator, children’s storybooks, Hugging Face, Google AI, Node.js, React, generative storytelling, text-to-speech AI sdp web application