Application-Driven Honey Selection via Principal Component Analysis-Aided Optimisation

Honey, a natural sweetener produced by bees from floral nectar, is known for its diverse physicochemical properties found in various applications (e.g., medical, dietary). The clinical utilisation of honey highlights the importance of antimicrobial and antioxidant benefits in various medical treatments. The primary factor contributing to the consumer purchasing behaviour of honey for dietary needs is mainly influenced by taste, texture, and colour. This challenged the honey selection process, ensuring the suitability and quality of honey in diverse contexts. This work focuses on addressing the multi-varied honey selection problem utilising Multi-Criteria Decision-Making (MCDM) analysis to identify key factors contributing to the selection of honey attributes based on the intended use. This provides a systematic and objective means of selecting honey varieties for clinical and dietary applications. The model favours the New Zealand Apis Manuka honey (Manuka2) for consumer dietary attributed to its high sugar concentration (33.24 wt%) and low acidity (4.04 mEq/kg) contributing to the desirable sweet and non-acidic flavour. The raw stingless bee honey (SBHR5) from Sarawak local bee farm exhibits impressive antimicrobial properties such as high free acidity (19.19 mEq/kg), DPPH assay (908.21 mg/mL), and total phenolic content (384.27 mg GAE/kg dwb) desirable for clinical use. This work explores the potential of utilising Principal Component Analysis (PCA)-aided optimisation for honey selection and its significance in identifying honey varieties with desired characteristics.