Akademik Veri Yönetim Sistemi
Journal of the Faculty of Engineering and Architecture of Gazi University, cilt.40, sa.3, ss.1659-1672, 2025 (SCI-Expanded)
Linguistic summarization, which has garnered significant attention in recent years, facilitates the derivation of human-understandable insights from vast amounts of data. One critical aspect of linguistic summarization involves determining the truth degree that reflects the extent to which the underlying data has been appropriately represented. In the extant literature, the numerical values of fuzzy sets employed to calculate truth degrees have been generated using the uniform partitioning method, which neglects the intervals of data concentration. To address this limitation, the present study proposes the Interval Type-2 Fuzzy C-Means (IT2FCM) partitioning method, which distributes fuzzy sets, taking into account the intervals of data density. By using the proposed method, interval type-2 fuzzy sets are created, and the truth degrees of linguistic summaries are calculated using fuzzy cardinality-based probability and possibility based approaches. The proposed approach is explicated step by step, and its outcomes are compared to those of the uniform partitioning method used in prior research. To evaluate the efficacy of the proposed IT2FCM partitioning approach, we apply it to financial time series covering the last decade of three stocks traded on the Borsa İstanbul (BIST). The variance of the obtained truth degrees is lower than that in equal partitioning studies, which indicates that the proposed method produces more stable results.