INFORMATION SYSTEM FOR RECOMMENDATIONS OF COSMETIC SKIN CARE PRODUCTS

Abstract

The growth of the range of skin care cosmetics with active components, as well as the possibility of their purchase without the appointment of a cosmetologist, causes the problem of the appearance of negative consequences of their use. The selection of cosmetic products is a task that can be solved using machine learning methods, such as neural networks. The article is devoted to the issue of improving the information support for the selection of cosmetic products by developing a recommendation information system. A study of the subject area was conducted and a FNN neural network of was chosen as a model for providing recommendations. Neural network models were built to select the most suitable component of a cosmetic product depending on the skin type and problem, as well as recommendations for cosmetic products. The results of computational experiments showed a high accuracy of the models of over 90%. This indicates that the introduction of machine learning into the process of forming recommendations allows for a high-quality and accurate selection of the desired cosmetic product. An integrated information system for providing recommendations has been developed, which consists of a client mobile interface, as well as software modules for providing recommendations written in Python. The interaction between the components of the information system is carried out through the API interface. The user in the mobile application can enter information about skin type, dermatological problems and receive appropriate recommendations for skin care. Using the developed information system when selecting care cosmetics will allow to avoid the use of products with incompatible components and avoid negative consequences for skin health.