Research And Implementation Of Recommendation Algorithm Based On Infant Growth Model

With the rapid development of the Internet and information technology,various industries have transformed into the Internet and made great progress.In many industries,the “Internet +” mother and child have been more prosperous than ever with the national policies of recent years and the family’s emphasis on parenting.As parents,people can learn about children’s growth through the Internet,find like-minded friends or infants,and explore or learn about baby-raising experience.The development of“Internet +” maternal and child industry has provided great convenience for users.However,with the development of Internet information,maternal and child users will inevitably face a lot of spam and meaningless data,so-called Information overload problem.Especially with the increasing youthfulness of women of childbearing age,the Internet has become a part of the lives of most parents.They are also more inclined to use the Internet to help them understand their children’s growth and get more scientific and reliable information that meets their needs.Therefore,the problem of information overload is more prominent.Based on this,this paper constructs a child growth prediction model based on multiple linear regression,and then studies and designs a collaborative filtering algorithm that integrates children’s growth information according to children’s growth information.Finally,using the proposed model and algorithm design to achieve a recommendation system for infant growth analysis.The main research contents are as follows:(1)Child growth prediction model based on multiple linear regression.Because of the limitations of using cross-sectional survey data to analyze the growth characteristics of children,longitudinal tracking data was used to construct a multiple linear regression prediction model for children’s height growth.Firstly,from the perspective of individualization,considering the growth and development characteristics of children,through the correlation analysis of gender,age and birth length,it is determined whether there is correlation with children’s height;secondly,with the help of SPSS(Statistical Product and Service Solutions)software,a multiple linear regression prediction model of children’s height was established,and the model was tested.The results show that the model has a good goodness of fit and has certain application value.(2)A collaborative filtering algorithm that integrates child growth information.For the traditionalcollaborative filtering technology,only the user and project problems are considered when generating recommendations,and a collaborative filtering algorithm that integrates children’s growth information is proposed.Because children’s information is a unique feature in the “Internet +” maternal and child industry,the traditional collaborative filtering algorithm does not consider children’s information,therefore,the algorithm first calculates child similarity based on children’s attributes and growth information;secondly,user attribute information is used to calculate user similarity;thirdly,the child similarity is weighted with the user similarity to calculate the total similarity of the user;finally,the user is recommended according to the total similarity of the user.Experimental results show that the algorithm can improve the accuracy of recommendation to a certain extent.(3)Design and implement a recommendation system for infant growth analysis.The system adopts C/S architecture.Through the analysis of system business and functional requirements,the system is divided into four modules: user,child,growth record and recommendation.The function of four modules is realized.The regression child growth prediction model is applied to the growth record module to provide users with child growth analysis,which is convenient for users to understand children’s growth status at any time.The recommendation module provides a personalized recommendation service by using a collaborative filtering algorithm that integrates child growth information,and presents the results to the user in a visual form.