With the development of Internet,big data and cloud computing technology,the degree of education informatization is deepening,and the digital education resources are becoming increasingly huge.The digital education industry has a good foundation.As an emerging mode in the Internet economy,crowdfunding and innovation mode is flexible,low threshold and fast spreading,which is conducive to the development and promotion of new technologies.However,due to the wide source,huge quantity,different structure,scattered and redundant content of educational resources,it is difficult for learners to obtain effective resources quickly.In order to strengthen the integration of digital education resources,improve the fusion level of information resources,and build a highquality digital education resource database,it is necessary to effectively align the heterogeneous knowledge bases from different sources.This dissertation studies the digital resource fusion technology under the crowdfunding and crowdfunding mode,proposes effective entity alignment methods for the fusion of encyclopedia knowledge base and open knowledge map,and designs and implements the evaluation system of educational resource fusion on this basis.The main contents of this dissertation are as follows1)This dissertation proposes a multi-source knowledge base index algorithm based on Bert model.This method constructs entity features and indexes through Bert model,which effectively improves the efficiency and accuracy of entity alignment between heterogeneous encyclopedia knowledge bases.2)This dissertation proposes an entity alignment method based on multi feature fusion.This method uses Bert and GCN to obtain the structure,entity name and attribute features of entities,uses sorting learning algorithm to fuse multiple features,and expands the training set through iterative training framework,which can effectively improve the effect of entity alignment between heterogeneous knowledge maps.3)Aiming at the educational resources in the process of fusion,the method and system of fusion data quality evaluation are designed.Get the education resources uploaded by users to be integrated or integrated through the network,evaluate the redundancy and lack of information of resources,and help users improve the quality of education resources. |