To predict the sensory evaluation scores based on the parameters determined by a texture analyzer, the raw rice samples were prepared according to the national standard GB1354-2009, and the cooked rice samples were prepared in large amounts and evaluated according to GB/T15682-2008. In this study, 60 lots of Japonica rice (round-grained) samples and 60 lots of Indian rice (long-grained) samples were collected from different regions. The palatability of cooked rice as an objective quality index was used as the independent variable, the subjective sensory evaluation score was obtained according to the national standard GB/T15682-2008 as the dependent variable, to establish a linear model between the textural parameters and the sensory evaluation score of the same rice by the partial least squares method. There were significant differences among the different rice samples in the overall sensory evaluation scores and textural parameters. But a linear correlation between the textural parameters and the sensory evaluation scores was found for the same rice. The equation for the obtained linear regression model has a minimum fitting degree of 95.86%. Therefore, the measurements of textural parameters using the texture analyzer allow an accurate prediction of the sensory evaluation scores of cooked rice samples.
