Information retrieval (IR) is a field in computer science that focuses on searching and retrieving relevant information from large-scale data sets. With the development of technology and the explosion of digital information, the need for efficient and accurate IR systems is increasing. This study aims to examine various IR models, both classical and modern, and the supporting techniques used in the information retrieval process. Classical models such as the Boolean Model, Vector Space Model (VSM), and BM25 are the initial foundations of IR systems, while modern neural network-based approaches such as DSSM, DPR, and Retrieval-Augmented Generation (RAG) offer more semantic and contextual search performance. In addition, basic techniques such as indexing and tokenization, as well as advanced techniques such as query expansion and relevance feedback, are also discussed, which also increase the effectiveness of the system. The performance evaluation of the IR system is carried out using various metrics such as precision, recall, F1-score, MAP, and NDCG. The results of this study indicate that the combination of the right IR models and techniques can produce an information retrieval system that is more relevant, efficient, and adaptive to the needs of modern users.
Copyrights © 2025