This paper aims to review existing techniques for identifying fragrances and digital representation and discuss the gap between fragrance detection and structured digital fragrance representation. This paper proposed the digital fragrance fingerprint (DFF) as a conceptual framework for structured digital fragrance representation using a literature review of existing works.
A systematic review of literature as determined by the preferred reporting items for systematic reviews and meta-analyses guidelines was conducted, and a total of 38 reference papers were studied. This review included gas chromatography-mass spectrometry (GC-MS), electronic nose (E-nose) technologies, nanomaterial-based sensors, signal processing, feature extraction and artificial intelligence (AI). Findings are analyzed to determine the potential of existing methodologies in the proposed DFF conceptual framework.
The chemical composition is accurately provided by GC-MS, while the quick recognition of fragrance is enabled by E-nose. Moreover, nanomaterials allow for improving sensors, and feature extraction combined with AI contributes to analyzing and classifying sensor data. Yet, limited studies provide structured digital fragrance information. Based on the literature findings, a conceptual DFF framework is proposed.
The proposed DFF is an approach to structuring the output of the process of fragrance detection, signal processing, feature extraction and AI analysis into a digital representation. This proposal should be considered a fundamental informational layer rather than a definitive standard for representation and should serve as a starting point for further research in this area.
