Experiment with Text Summarization as a Positive Hierarchical Fuzzy Logic Ranking Indicator for Domain Specific Retrieval of Malay Translated Hadith

Expert Judgment Z-Numbers as a Ranking Indicator for Hierarchical Fuzzy Logic System

“…In term of the overall queries, the proposed system, with the additional new indicator of Expert Judgment Z-Numbers, may bring the improvement to the original BM25 ranking function, by yielding better results on 26 queries, on all evaluation metrics that are measured in this research such as P@10, %no measures and MAP, and has achieved the better results in 28 queries on P@10 alone, compared to the BM25 original score, that only yield better result in two queries of all evaluation metrics and four queries on the [21], the proposed system has a better result on five topics and 26 queries. With the additional of the Expert Judgment with Z-Numbers in this research, the results are improved when the proposed system yields better results on seven topics and 26 queries compared to the other model.…”

“…The system is designed based on the integration of the researchers' two previous works which are, Domain specific concept ontologies and text summarization as hierarchical fuzzy logic ranking indicator on Malay text corpus in [2], and Positive Rate of Hadith in [21], with the new additional Expert Judgment with Z-Numbers that will be the main contribution in this article. The details of the design of this research will be explained as follows.…”

“…One advantage of Hierarchical Fuzzy Logic System is its capability to decrease the amount of fuzzy rules in the complete system, by separating the input in different Fuzzy Logic Controller (FLC), this approach also will decrease the computational time and ensuring the systems' robustness and its efficiency, moreover proven to capture and manipulate the expert knowledge and input [22]. In this article, the researchers once again use the design of a Hierarchical Fuzzy Logic System, which has been done in the researchers' previous work in [2,21]. In [2], the researchers have explained the operation FLC1 and FLC2, which is in the Fuzzy Logic Controller (FLC1) of Mamdani Method, consist of two inputs such as such as Ontology BM25 Score and Fabrication Rate.…”

“…For the FLC2, it consists of two inputs such as the output of FLC1 and Shia Rate. In [21], the researchers have explained the continuous design by adding the new FLC3, which consists of two inputs such as the output of FLC2 and Positive Rate of Hadith. In this article, the researchers have added a new FLC4, that run after FLC3 process that also consists of two inputs, such as the output of FLC3 and Expert Judgment Z-Numbers, and yields the output as the Final Ranking Score of the document shown in Figure 2.…”

“…In this article, the researchers explain the linguistic variables as shows in Table 1 and Table 2, formulation and calculation of new Expert Judgment Z-numbers respectively that will be used in FLC4. The details of the linguistic variables, the formulation and the calculation of FLC1, FLC2, and FLC3 which can be seen in the researchers' previous works in [2,21]. In FLC4, the model uses one input which is the Expert Judgment Z-Numbers and fouroutput values of the Final Score Ranking which consists of three triangular membership functions.…”