Ritthiruangdej Pitiporn, Suwonsichon Thongchai
Department of Product Development, Faculty of Agro-Industry, Kasetsart University, Bangkok, Thailand.
Anal Sci. 2007 Jul;23(7):809-14. doi: 10.2116/analsci.23.809.
Twenty Thai commercial fish sauces were characterized by sensory descriptive analysis and near-infrared (NIR) spectroscopy. The main objectives were i) to investigate the relationships between sensory attributes and NIR spectra of samples and ii) to characterize the sensory characteristics of fish sauces based on NIR data. A generic descriptive analysis with 12 trained panels was used to characterize the sensory attributes. These attributes consisted of 15 descriptors: brown color, 5 aromatics (sweet, caramelized, fermented, fishy, and musty), 4 tastes (sweet, salty, bitter, and umami), 3 aftertastes (sweet, salty and bitter) and 2 flavors (caramelized and fishy). The results showed that Thai fish sauce samples exhibited significant differences in all of sensory attribute values (p < 0.05). NIR transflectance spectra were obtained from 1100 to 2500 nm. Prior to investigation of the relationships between sensory attributes and NIR spectra, principal component analysis (PCA) was applied to reduce the dimensionality of the spectral data from 622 wavelengths to two uncorrelated components (NIR1 and NIR2) which explained 92 and 7% of the total variation, respectively. NIR1 was highly correlated with the wavelength regions of 1100 - 1544, 1774 - 2062, 2092 - 2308, and 2358 - 2440 nm, while NIR2 was highly correlated with the wavelength regions of 1742 - 1764, 2066 - 2088, and 2312 - 2354 nm. Subsequently, the relationships among these two components and all sensory attributes were also investigated by PCA. The results showed that the first three principal components (PCs) named as fishy flavor component (PC1), sweet component (PC2) and bitterness component (PC3), respectively, explained a total of 66.86% of the variation. NIR1 was mainly correlated to the sensory attributes of fishy aromatic, fishy flavor and sweet aftertaste on PC1. In addition, the PCA using only the factor loadings of NIR1 and NIR2 could be used to classify samples into three groups which showed high, medium and low degrees of fishy aromatic, fishy flavor and sweet aftertaste.
通过感官描述分析和近红外（NIR）光谱法对20种泰国商业鱼露进行了表征。主要目标是：i）研究样品的感官属性与近红外光谱之间的关系；ii）基于近红外数据表征鱼露的感官特征。使用由12名经过培训的小组成员进行的通用描述分析来表征感官属性。这些属性包括15个描述词：棕色、5种香气（甜、焦糖味、发酵味、鱼腥味和霉味）、4种味道（甜、咸、苦和鲜味）、3种余味（甜、咸和苦）以及2种风味（焦糖味和鱼腥味）。结果表明，泰国鱼露样品在所有感官属性值上均表现出显著差异（p < 0.05）。在1100至2500 nm范围内获得了近红外透反射光谱。在研究感官属性与近红外光谱之间的关系之前，应用主成分分析（PCA）将光谱数据的维度从622个波长减少到两个不相关的成分（NIR1和NIR2），它们分别解释了总变异的92%和7%。NIR1与1100 - 1544、1774 - 2062、2092 - 2308和2358 - 2440 nm波长区域高度相关，而NIR2与1742 - 1764、2066 - 2088和2312 - 2354 nm波长区域高度相关。随后，还通过PCA研究了这两个成分与所有感官属性之间的关系。结果表明，前三个主成分（PC）分别命名为鱼腥味成分（PC1）、甜味成分（PC2）和苦味成分（PC3），共解释了66.86%的变异。在PC1上，NIR1主要与鱼腥味香气、鱼腥味和甜余味的感官属性相关。此外，仅使用NIR1和NIR2的因子载荷进行的PCA可用于将样品分为三组，分别显示出高、中、低程度的鱼腥味香气、鱼腥味和甜余味。
