Design And Implementation Of Medical Cosmetology Application Based On Personalized Recommendation

With the improvement of living standards,people have higher requirements for the quality of life.Especially for some people who pay attention to their appearance,it is an exciting thing to change their appearance.Under the trend of vigorous development of medical cosmetology,people’s understanding of this field of in China is still at the stage of high prices and backward technology.Therefore,the medical cosmetology platform in this paper innovates the model of e-commerce+community under the traditional e-commerce model.It provides a docking platform for medical cosmetology institutions,physicians and consumers to provide more thoughtful preoperative services and postoperative services,as well as postoperative experience sharing for customers.At the same time,the communication and feedback between consumers can also effectively enhance consumers’ trust and consumption experience,which can change the situation of information asymmetry in the medical cosmetic industry,and help the industry to form a healthy development pattern.The writer is mainly responsible for the design and development of the community part of the system,so this paper will introduce the community part of the system in detail.The main function modules of the community part include beauty diary module,Q and A post module,friends module,my message module,diary recommendation module and personal center module.Personal center module is divided into diary management sub-module,question and answer management sub-module,album management sub-module,order management sub-module,membership center sub-module,task center sub-module and so on.The algorithm of diary recommendation module is mainly divided into recall layer algorithm and rank layer algorithm.The recall layer uses Collaborative Filtering Algorithm and Content-based Recommendation algorithm to get the recommendation list.The ranking layer uses the Gradient Boosting Decision Tree and Logistic Regression model to predict the click-through rate of the recommended list to display it for users in the order of click-through rate from high to low.In Collaborative Filtering Algorithm,considering the influence of sparsity of traditional user scoring matrix on recommendation results,this paper incorporates clustering idea and Slope One algorithm to predict and fill the scoring matrix.In the process of mining user clustering,the consideration of user preferences is included.Considering the need to maintain two sets of codes and waste development manpower in the development of Android and IOS systems,the medical cosmetology platform in this paper chooses Ionic framework to realize mobile cross-platform,which reduces the maintenance cost and maintains the original features of the mobile terminal.The server side is developed with SSM framework.In the recommendation module,the recommendation system is divided into data loading layer,model training layer and service layer,using big data platform and related technology to calculate and store data.At the same time,Redis and Kafka are used to achieve high real-time performance.