Research And Implementation Of Thematic Database System For Agricultural Groups In Recent China

With the rapid development of information technology,more and more paper-based resources are transformed into digital resources for storage.The management method of ancient literature resources is gradually changed from manual management to information management based on database.Digital storage and management can not only improve the protection and sharing of document resources,but also improve the efficiency of searching for literature resources.This dissertation implements the thematic database system of recent agricultural groups in China,and this thesis studies and implements the recommendation methods of digital ancient literature resources.The project relies on the research results of the research group.We have scientifically sorted out and digitized the ancient literature resources,long-compiled materials and the results of the chronological compilation.Base on the digital resources,we implement the first large-scale thematic database system of recent agricultural groups in China and establish a three-dimensional retrieval subsystem.For facilitate user retrieval,this dissertation propose a recommendation algorithm based on text embedding and user portraits.The main work is as follows:1)According to the bidding document of National Social Science Fund Project,this dissertation analyze the interaction between researchers and literature resources,and this thesis complete the requirement design,technical architecture selection,database design and key function design of the project.The main functions of the system include: foreground content display,background content management,three-dimensional retrieval,personalized recommendation.2)Based on Spring,Hibernate,FreeMarker and other technologies,the thematic database system of recent Chinese agricultural groups is constructed.The system provides an easy retrieval function,which is convenient for researchers of universities and research institutions,relevant personnel of government departments and ordinary users to search resources.3)A recommendation algorithm based on text embedding and user portrait is proposed to improve the pertinence of recommending literature resources for users and improve the service quality.TF-IDF method is used to embed text into vector space.Cookie technology is used to identify anonymous users.User portrait is generated according to the user's browsing history.Euclidean distance is used to measure the similarity between user portrait and text vectors.In vector space,the algorithm recommends the first N unseen resources closest to the user's portrait to the user.The algorithm ensures that the recommended literature resources are related to the user's interest in the content and improves the user's experience.