An analytical review of optimization techniques in information retrieval for enhanced decision support

As digital content continues to evolve, enhancing Information Retrieval (IR) systems is crucial to improve their performance, relevance, and ability to handle increasingly large datasets. This systematic review examines current advancements in IR optimization strategies, with a focus on the paradigm shift from traditional heuristics to AI-driven models. Following the PRISMA 2020 guidelines, an exhaustive literature search was conducted for publications between January 2013 and June 2025, employing a hybrid screening approach that combined Natural Language Processing (NLP) automated filtering with manual expert review. Our findings underscore the growing importance of hybrid models that leverage deep learning, particularly transformer architectures, for tasks such as personalization and relevance feedback, which have demonstrated significant performance improvements. However, significant challenges such as algorithmic bias, computational complexity, and domain-specificity impede wider implementation. This review provides a comprehensive roadmap of the IR optimization landscape, identifies persistent ethical challenges, and explores emerging research frontiers, such as quantum IR and generative models, thereby offering actionable insights for both researchers and practitioners in the decision analytics field. • Apply a structured review methodology to examine optimization strategies in information retrieval systems. • Analyze 32 empirical studies from a decade of research on retrieval efficiency and relevance. • Identify major techniques such as feedback, suggestion, and personalization to enhance search outcomes. • Examine how retrieval strategies improve accuracy, recall, and user satisfaction across applications. • Highlight emerging challenges in scalability, ethics, and domain-specific optimization practices.