Assigning Values to Uncertainty Functions of Linguistic Terms for Evaluation Scale Construction: Process Capability Analysis Perspective

Computing with Words (CWW) is a foundational concept in computational intelligence that aims to process and manipulate linguistic information generated from human reasoning and decision-making. With its providing, the generated models have ability to handle uncertainty and imprecision in computational systems, by increasing reflection rate of natural and human-like interactions and decision-making processes in the outputs. One of these computational systems, process capability analysis (PCA), is a tool in statistical process control aims to provide process variability with different analyses. The key concept in it is accuracy of the analyses, which is directly related to the precision of the specification limits. When the input data involves uncertainty, the crucial step is the processing of the input data without loss of information. By extending PCA with fuzzy logic, it aims to reflect the input data without loss of information by assigning membership functions. Moreover, since the system may involve different types of uncertainty, this paper proposes a novel evaluation scale construction for linguistic terms under uncertainty by involving fuzzy z-numbers, which enables to represent not only the preciseness of the measurement but also the reliability of the measurement system. An application is also carried out for the illustration and applicability of the proposed model. Based on the results, it is essential to acknowledge potential limitations, such as the generalizability of the approach across different contexts and the need for empirical validation in diverse scenarios.