Application of Fuzzy Tahani and TOPSIS Methods for Personalizing the Coffee Brewing Process According to Consumer Taste Preferences

Abstract

Coffee brewing quality is determined by parameters such as bean density, water temperature, grind size, and brew ratio, each influencing the final brew strength classified as light, medium, or strong. Variations in these factors make it difficult to achieve consistent extraction outcomes through sensory evaluation alone. To address this limitation, a web-based decision support system was developed by integrating the Fuzzy Tahani inference model with the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) ranking method. The system evaluates 1,309 parameter combinations derived from empirical ranges of density (350–450 kg/m³), temperature (90–96 °C), grind size (500–1300 µm), and brew ratio (1:12–1:18). Three quantitative suitability functions measure density–temperature, grind–strength, and ratio–strength alignment, while a fourth quantifies fuzzy rule compatibility. These functions are aggregated using TOPSIS to rank valid alternatives and determine the optimal configurations for each brew-strength category. Validation through controlled pour-over experiments confirms the computational model: Total Dissolved Solids (TDS) decrease proportionally with increasing bean density, whereas Extraction Yield (EY) remains stable within the ideal 18–22 % range as defined by the Coffee Brewing Control Chart. The integration of Fuzzy Tahani and TOPSIS bridges linguistic reasoning with quantitative evaluation, yielding reproducible recommendations for consistent and precise coffee-brewing control.