Akademik Veri Yönetim Sistemi
Tehnicki Vjesnik, cilt.33, sa.1, ss.294-304, 2026 (SCI-Expanded, Scopus)
Recently, big data has become increasingly important in the fields of scientific research and application. However, due to the characteristic features of big data such as high volume, velocity, variety, variability, value, and complexity, processing it with traditional analysis methods is quite a challenging process. In this context, frameworks like MapReduce are commonly used in the modeling of big data and in parallel and distributed data processing techniques. In this study, it is aimed to use fuzzy functions based on the fuzzy c-means (FCM) algorithm under the MapReduce architecture for modeling systems based on large data sets. In the study, it is explained in detail how the FCM algorithm is parallelized in the mapping phase; subsequently, it is demonstrated how the data is reduced in the reduce phase and how the fuzzy functions are derived. The proposed approach demonstrates the effectiveness of fuzzy functions within the MapReduce framework in modeling systems based on various large datasets. Additionally, the success of the methodology has been thoroughly discussed through the evaluation of the obtained fuzzy functions and performance analysis.