Design And Implementation Of Health Information Service Recommendation Software Intelligent Pension System

In recent years,the aging of China’s population has become increasingly serious,and the emergence of intelligent pension has effectively alleviated the pension problems of the current society.The elderly are prone to disease problems due to low immunity,and the accumulation of health information knowledge can reduce the probability of illness to a certain extent,so health information service software is particularly important.However,in the current market,most health information service software only recommends health information based on user clicks and release time,and the recommendation results lack accuracy and personalization.In addition,massive amounts of data make it difficult for the elderly to obtain accurate and valuable health information.In response to these problems,this thesis takes the recommendation algorithm as the core to design and implement a health information service recommendation software for intelligent recommendation based on the historical behavior and social relationships of the elderly.The main work of this thesis is as follows:1.This thesis investigates the background and the current research status at home and abroad of the subject,formulates the expected goals of the elderly health information service recommendation software,and combines the needs of the elderly to propose an overall design plan based on the B/S development framework;2.This thesis analyzes and studies the elderly health information recommendation algorithm that combines user preferences and trust relationships.In order to improve the accuracy of the recommendation algorithm,this thesis mines the user preferences and trust relationships of the elderly from the historical behavior data and social relationship data of the elderly.Firstly,this thesis obtains the user-label-information three-dimensional relationship based on historical behavior data,and calculates the correlation and the information content respectively of the split user-tag two-dimensional relationship and the tag-information two-dimensional relationship to obtain the user’s preference model.Through TF-IUF method calculates the importance of information to users and obtains the information weight.By introducing the time factor,the time factor similarity coefficient is proposed,and the information weight and the time factor similarity coefficient are revised to the user’s preference similarity.Secondly,this thesis constructs a social network based on the user’s social relationship,calculates the direct trust through the frequency of interaction and social influence,selects the long path with high trust as the effective trust path,and the average trust of multiple effective trust paths is used as the user’s indirect trust.The combination of the direct trust and the indirect trust obtains the social trust.Finally,this thesis adopts the method of dynamic fusion to adjust the weight of user preference similarity and social trust to obtain the hybrid similarity between users.3.According to the software engineering specification process,this thesis describes the design of the entire software and each functional module in detail,and tests the functions and performance of the software by formulating test cases.The test results show that the recommended software for the elderly health information service meets the expected function and performance requirements,and achieves the expected goals.