Studija slučaja

https://doi.org/10.32985/ijeces.13.10.11

Information retrieval approaches: A comparative study

Assmaa Moutaoukkil ; IRF-SIC Laboratory, Department of Computer Science, Faculty of science, Ibn Zohr University, Agadir, Morocco
Ali Idarrou ; IRF-SIC Laboratory, Department of Computer Science, Faculty of Science, Ibn Zohr University, Agadir, Morocco
Imane Belahyane ; IRF-SIC Laboratory, Department of Computer Science, Faculty of Science, Ibn Zohr University, Agadir, Morocco

Sažetak

The area of information retrieval (IR) has taken on increasing importance in recent years. This field is now of interest to large communities in several application domains (security, medicine, aeronautics, etc.). IR studies find relevant information from the semi-unstructured type of data. As the information resources generated after the search can be extensive in quantity and different in quality, it is essential to rank these results according to the degree of relevance. This paper focuses on text information retrieval (TIR) and emphasizes the importance of each IR approach. This study presents insightful aspects of TIR and provides a comparative study between some proposed approaches and models. Each model offers IR advantages and suffers from several limitations.

Ključne riječi

information retrieval; NLP; graph; TF-IDF; document classification; KNN; fuzzy K-means; deep learning; BERT; transformers;

Hrčak ID:

290496

URI

https://hrcak.srce.hr/290496

Datum izdavanja:

21.12.2022.

Posjeta: 634 *