Avaliação sensorial de manga (Mangifera indica L. cv. Tommy Atkins) osmoticamente desidratada por meio de analise descritiva quantitativa e redes neurais artificiais.

Abstract

The production of intermediate moisture foods have been shown especially for developing countries, lince it requires simple technologies, generating quite stable products, in addition to energy savings and lower capital investment. The food industry has sought to identify and meet the desires of consumers toward their products. This is accomplished with the study of sensory analysis has been an important tool in this process. It is a science that aims mainly to study the perceptions, sensations and reactions of consumers about the characteristics of the products, including its acceptance or rejection. The combination of methods (osmosis and drying) has been identiüled as safe and economical alternative for storing various food products. This research enables the physical-chemical, microbiological and application of new techniques in order to improve production and storage and facilitate the study of acceptance by consumers. The fruits of mango cultivar Tommy Atkins were hand picked and selected according to the maturation stage, and went through severas processing steps and then were analyzed physico-chemical, microbiological and sensory. During the study it was found that the drying process resulted in an increase in the percentage of soluble solids and total sugars due to the reduction of the moisture content. With regard to microbiological tests, the results demonstrated in this study reflect the hygienic conditions employed in the preparation and maintenance of quality. The sensory acceptance of osmotically dehydrated manga as CUBOlh presented better resulta than the samples toothpick and CUBE This happened due to the sample stock present itself firmer and drought which caused a lower uptake of the product to be sampled. The use of quantitative descriptive analysis provided relevant information that describe resulta showing significant differences for all attributes between the 3 samples tested, except for the sour taste that showed similarity to the treatments CUBOlh and stock to the same attribute. Therefore, it was concluded that the study of Sensory Analysis with Artificial Neuras Networks will collaborate with technological advances during the surveys conducted by industry and by academics in search of products and increasingly elaborate exactly according to customer preference, But more study is necessary to increasingly improve the method.