Enhancing Multimodal Information Retrieval Strategies to Improve Access and Discovery in Digital Library Services

Abstract

The development of digital libraries has simplified the storage and retrieval of videos, pictures, and text files. However, traditional keyword searching, which continues to be the primary method of digital libraries, still suffers from issues concerning relevance, accessibility, and adaptability. The primary aim of this manuscript is to address these problems by creating modern and sophisticated algorithms for Multimodal Information Retrieval (MIR). The system uses artificial intelligence, machine learning, and natural language processing to integrate personalization and user-centric frameworks, profile-driven semantic search, and tailored user interfaces to enhance system responses. Relevant recent studies show that automated and intelligent metadata assignment, word sense disambiguation, and named entity recognition improve precision, recall, and F1-score in retrieval tasks significantly, especially for multilingual and multi-format datasets. In addition, the study broadens the application of FAIR (Findable, Accessible, Interoperable, and Reusable) principles, focusing on AI knowledge graphs, AI recommendation systems, and contextualized learning systems that enhance discoverability in digital libraries. The research concluded that these systems improve precision and, at the same time, grow user participation, engagement, inclusiveness, and scalability beyond the physical bounds of libraries, turning them into dynamic information systems in Open Science.