A Machine Learning Approach Investigating Consumers' Familiarity with and Involvement in the Just Noticeable Color...

Area of Science:

• Food Science
• Sensory Science
• Colorimetry

Background:

• Visual color perception is crucial for food quality assessment.
• Instrumental color measurements provide objective data but require interpretation in relation to human perception.
• Consumer perception of food color is influenced by various factors, including familiarity and involvement.

Purpose of the Study:

• To determine the relationship between instrumental color and visual color perception of dry-cured ham.
• To establish the Just Noticeable Color Difference (JNCD) for dry-cured ham color.
• To investigate the influence of consumer involvement and familiarity on color perception and JNCD.

Main Methods:

• Instrumental color analysis of dry-cured ham slices.
• Consumer surveys assessing color scoring, matching, involvement, and familiarity.
• Clustering consumers based on involvement levels.
• Calculation of JNCD for different consumer clusters.
• Application of interpretable machine learning to correlate visual and instrumental color data.

Main Results:

• A JNCD of ΔEab* = 6.2 was determined for dry-cured ham color.
• JNCD was found to be lower for younger consumers.
• Consumer involvement significantly influenced color perception and JNCD.
• Machine learning models incorporating psychographic data outperformed multiple linear regression.
• L* and hab were identified as the most influential color variables in the machine learning model.

Conclusions:

• Consumer involvement and familiarity significantly impact the perception of dry-cured ham color.
• Objective color measurements (L*, hab) combined with machine learning can effectively model visual color perception.
• The established JNCD provides a benchmark for quality control and product development in the dry-cured ham industry.