Texture prediction and influencing factor analysis of gel-type foods using machine learning based on compression and penetration data

Puntos clave

• This study aims to connect the mechanical properties of gel foods with sensory perceptions using machine learning.
• Constructed a machine learning model based on sensory evaluation and compression/penetration test data
• Analyzed specific onomatopoeic expressions related to texture characteristics
• Investigated the relationship between mechanical properties and sensory impressions
• Identified correlation between specific onomatopoeia and compression/penetration characteristics
• Demonstrated that texture perceptions can be linked to mechanical measurements
• Established a foundation for using sensory evaluations as scientific indicators in food development