Research On Personalized Recommendation Of Library Resources Based On User Profile

The current resource recommendation of the library management system of the university library is to push the content without distinction,which is in contradiction with the needs of diversified readers;at the same time,the construction of the smart campus is in the process of middle school work,personnel,educational affairs,clubs,activity centers,library management,and retrieval.A large amount of data has been accumulated in many systems,and the value of the data needs to be explored.In order to solve these problems,learn from the experience in the business field,apply user profile technology to the library,carry out reader portrait research in a smart campus environment,master the basic characteristics,interest preferences and potential needs of different readers,in order to realize books and lectures and other libraries Personalized recommendation of resources,thereby optimizing library information services,enhancing user experience,and increasing resource utilization.The article proposes the process of constructing reader portraits in a smart campus environment,including data collection and preprocessing,user interest modeling,and user profile display.By sharing the data of various information systems in the school data center,collecting user basic information and borrowing and returning books,searching,participating in academic reports,participating in topic lectures,message board messages,comments and other behavioral data,the data is integrated and cleaned to establish basic information and learning The user profile framework of five aspects: needs,book borrowing,digital resources,and other hobbies,determines the generation rules and storage structure of user profiles,establishes user subject preference models and topic preference models,and uses statistical analysis,text mining and other methods to refine users Tags and calculate tag weights to generate reader portraits.The article also designed a resource recommendation system to complete the personalized recommendation of resources.It mainly includes four functional modules: user profile calculation,resource feature extraction,recommendation result generation,and recommendation result transmission.The functions and specific implementation methods of each module are carried out.Discuss in detail.Finally,an empirical study was carried out with students majoring in educational technology as an example.The recommendation effect was evaluated through user satisfaction survey questionnaires and comparison of borrowing before and after recommendation.The data showed that the recommendation results were highly targeted and accurate,indicating that the use of user profiles was used.The method of personalized recommendation is feasible and can provide reference for the research and practice of personalized recommendation of various resources.