Hassan Abdulsattar Ibrahim

The English Translation of the Quran Tafsir (ETQT) is essential to understanding and interpreting Allah’s words. Clustering is a common text mining technique for eliciting meaningful knowledge from a text collection. It is commonly used when the selected datasets lack typical ground truths. To the best of our knowledge, no study has evaluated and benchmarked ETQTs to select the most comprehensive and appropriate one. The process of evaluating and benchmarking ETQTs falls under the multicriteria decision-making (MCDM) problem because of different issues, namely, multiple internal clustering validation criteria, data variation, and trade-offs between different criteria. The fuzzy decision by opinion score method (FDOSM) is one of the most recommended MCDM ranking methods in the literature to address the said issues. FDOSM has been extended under different fuzzy set (FS) environments to address issues of uncertainty and vagueness caused by expert feedback subjectivity. Although prior versions of FDOSM improved the uncertainty and vagueness issues, they remain open issues. Therefore, this paper extended FDOSM into the complex Pythagorean fuzzy decision by opinion score method (CPFDOSM) to evaluate and benchmark ETQTs. The proposed method consists of two main phases. The first phase formulates decision-matrix-based cluster algorithms and internal cluster validation criteria. The second phase (CPFDOSM development) prioritizes ETQTs and selects the optimum one. Data generation is performed on five different cluster algorithms and six internal cluster validation criteria using 16 ETQTs based on three decision-makers (DMs). Results show the following: (1) 6.25% of the individual decision-making results are identical among the three DMs, whereas 93.75% ([ ]) are different when [ ] and [ ]. When [ ]. In addition, T7 has consistent ranks (Rank[ ]=[ ]16) across all DMs, whereas T14 has consistent ranks (Rank[ ]=[ ]1) across all DMs when [ ]. (2) The results of the group decision-making reveal T14 is the most comprehensive and appropriate ETQT across all [ ] values. Objective validation and comparison analysis show the proposed method is ranked systematically.