Unveiling the Secrets of Black Tea: Artificial Intelligence in the Analysis of Sensory and Bioactive Parameters

Favorite this paper
How to cite this paper?
Details
  • Presentation type: Poster
  • Track: Sensory characterization of food, consumer-food relation – (CS)
  • Keywords: Antioxidants; Black tea; Machine learning;
  • 1 UFPEL

Unveiling the Secrets of Black Tea: Artificial Intelligence in the Analysis of Sensory and Bioactive Parameters

Layla Damé Macedo

Federal University of Pelotas

Abstract

Black tea, one of the most popular beverages globally, stands out for its functional properties and health benefits. Its processing involves the full oxidation of Camellia sinensis leaves, resulting in the formation of bioactive compounds like theaflavins and thearubigins, which are responsible for its dark color and intense flavor. Although not essential for nutrition, these compounds possess antioxidant and anti-inflammatory properties. The use of artificial intelligence (AI) in the food industry has grown, allowing for the optimization of sensory and physicochemical analyses, modeling consumer preferences, and controlling factors that affect the final product's quality. In this study, the aim was to use AI to analyze the physicochemical and sensory properties of black tea. First, sensory data were collected through acceptance tests, where one hundred and fifty participants evaluated color, appearance, aroma, and flavor using a 7-point hedonic scale. Then, physicochemical variables such as color (colorimetry), pH, and antioxidant determination methods (DPPH, ABTS, FRAP, and ORAC) were measured. The Random Forest model was applied to evaluate the variables, aiming to identify the most influential factors on the overall impression of the tea and determine the most effective method for measuring black tea's antioxidant capacity. The data were processed using Python programming language and the SHAP (SHapley Additive exPlanations) technique, along with the model's metrics (MAE and MSE) for result interpretation. The model showed that color, aroma, flavor, and appearance were the most relevant factors, with color being the most important. The model's performance was significant, with an MAE of 0.36, MSE of 0.12, and 88% accuracy. Additionally, the model demonstrated that the DPPH analysis was the most effective for determining antioxidants in black tea, with an MAE of 0.15 and 91% accuracy. The conclusion highlights that AI can optimize analyses, providing valuable insights to improve quality and meet consumer preferences.

Share your ideas or questions with the authors!

Did you know that the greatest stimulus in scientific and cultural development is curiosity? Leave your questions or suggestions to the author!

Sign in to interact

Have a question or suggestion? Share your feedback with the authors!