This study shows a methodology which correlates sensory perception and predictive simulations performed in a model based on artificial neural network (ANN), inputted with textural data. The study used 39 sunscreen samples available in Asia, North America, Latin America, and Europe. Experimental textural analyses considered spreadability, back extrusion, and tackiness and the results were inputted in the ANN model. Sensory tests with 325 panelists distributed in the four mentioned regions (approximately 80 panelist per region) were carried out and 14 attributes were evaluated. The correlation between the predictive simulations and sensory tests showed precision of 60-84% for the majority of attributes. Few attributes showed correlation in the range of 37-60%. This study shows as a practical application a model to estimate consumer acceptance of sunscreen products using ANN and experimental textural results. This tool proved to be reliable, being a competitive advantage in the development of new cosmetic emulsions.
Practical applicationsThe assessment of textural and sensory characteristics of cosmetic products is critical for their development and further acceptance by consumers. Ideally, the most accurate evaluation is the experimentation in use by screened consumers (panelists), but this test is equally time-consuming and expensive. In this scenario, this work presents as a practical application a tool to predict consumer acceptance of cosmetic emulsions based on an artificial neural network model inputted with textural data.