Research On Personalized Cuisine Recommendation Under Large Data Environment

With the rapid development of science and technology and rapid economic development,people's living standards have been improved gradually,the pace of life has been accelerated,and more and more people are going out to eat.However,"the people are eating the day","the body is the capital of the revolution",how to ensure the health of the people,and provide people with a reasonable and healthy diet,and Catering industry is being paid more and more attention,and the personalized service level of catering industry is also urgent to improve.In addition,smart phones have basically achieved one hand,and the development of e-commerce has gradually shifted from traditional desktop computers to mobile terminals.As the catering platform is located in the mobile terminal,the mobile terminal has also become a new battleground in the catering industry.It is one of the effective ways to improve the user experience by providing personalized recommended dishes for users.To this end,this paper builds a personalized cuisine recommendation system,through which the system can provide different users with the user needs of the dishes,is very suitable for the current hot network ordering environment.The personalized cuisine recommendation system in this paper consists of three parts,which are the feature extraction part,the user feature extraction part and the recommendation algorithm part.The feature extraction of the dish feature extracts the feature information of the dishes,the user feature part is used to extract the user's behavior information,and the algorithm is recommended to calculate the similarity between the dishes and the users.In this paper,the overall design of the system has been carried out,and the research on dishes and user feature extraction has been carried out,and the prototype has been realized.The main work is as follows:1.To extract the multi-dimensional features of the dishes,on the one hand,according to the characteristics of the quality of the dishes,on the other hand,according to the relationship between the dishes and the dishes,the characteristics of the dishes are found,and the characteristic index system of the dishes is constructed,and the relationship between the dishes and the dishes is set up.2.To extract the user's characteristics,first,according to the user's popular eating habits,the popular catering culture,and the user's own conditions,extract the dominant characteristics of the user's catering,and then combine the consumer's dietary consumption,excavate the recessive characteristics of the user's consumption preference,and build the user's characteristic system based on the analytic hierarchy process.3.Through similarity algorithm,according to the characteristics of dishes and user characteristics,the similarity between dishes and users is calculated,and the recommended results are generated.4.In summary,a prototype of personalized cuisine recommendation system is designed in this paper,and the prototype system is implemented based on the open source EasyRec recommendation engine.