Gelatine is frequently used as a food ingredient. However, Indonesia imports almost all of its gelatine, totaling 3990152 tons annually. Gelatine could be replaced with glucomannan compound which was found in porang tubers. However, it also contains calcium oxalate, which is harmful for the human body. In this study, calcium oxalate was first eliminated by the purification process using 10 % NaCl (w/w). Moreover, the microwave-assisted extraction method was used to extract the glucomannan compound by applying 300 W of microwave power with different extraction times (5, 10, 15, and 20 min) and different ethanol concentrations (60, 70, 80, and 96 %). Statistical analysis was used to optimize and identify significant parameters influencing the glucomannan concentration. The best conditions for glucomannan extraction were an extraction time of 10 min and an ethanol concentration of 80 % (v/v), resulting in a glucomannan yield of ≥96 %. Machine learning was successfully applied for data modelling using a Long Short-Term Memory block with an average R-square of 0.9772 (97.72 % accuracy) and an average MSE of 4.7719. Furthermore, physical and chemical characteristics of the extracted porang flour were accorded with SNI gelatine standards 06-3735 in 1995, which consisted of glucomannan (96.359 ± 1.164 %), calcium oxalate (0.009 ± 0.001 %), water (2.290 ± 0.986 %), ash (0.018 ± 0.002 %), fat (0.0235 ± 0.120 %), heavy metals (not identified), and pH (6.455 ± 0.191). Finally, the extracted glucomannan can be used as a potential regional substitute for gelatine production.
Keywords: Data modelling; Gelatine; Glucomannan; Microwave assisted extraction; Porang tubers.
© 2023 The Authors. Published by Elsevier Ltd.