Design And Implementation Of Knowledge Graph System Based On Heterogeneous Resources

The rapid development of Internet technology brings people great convenience,but also brings people inconvenience and challenges.Because of the rapid increase of data volume,people always find it more difficult to get the information they want.Knowledge Graph represents the information and relationship of all things in the world,Because of the association between these massive data,it can form high-quality,obvious and intuitive knowledge.This kind of Graph is not only convenient for people's understanding,but also more helpful for computer understanding and processing,so Knowledge Graph is helpful for people to get what they want and has become the focus of many scholars.However,there are many heterogeneous resources in the massive data.How to deal with heterogeneous resources and store them effectively in the Knowledge Graph system is a huge challenge.Based on the above background,a Knowledge Graph system based on heterogeneous resources is designed and implemented.First of all,the design process of Knowledge Graph system is introduced from the perspective of data storage,graph display and data analysis,and Data Schema in the implement of Knowledge Graph system is introduced briefly.Then it explains the function flow of the storage module,display module and analysis module.the storage module is implemented and the related operation of the storage module is defined by using the migration design and custom design of the Data Schema.The migration design reads different types of files,and the custom design is that the user defines model node.Finally,through Knowledge Extraction,information in massive text can be extracted and converted into knowledge,and Knowledge Extraction combine these two ways to store,so that heterogeneous and complex data processing can be realized.Using the idea of BS embedded in CS system,the display module of knowledge Graph is implemented,and the storage function and display function of the knowledge Graph system can also be carried out in the web page.The experiment is trained and tested in MSRA corpus.The experimental results show that Knowledge Graph system has a certain improvement in the accuracy and recall rate of extraction results,The F1 is about 88% and can be applied to the field of Chinese text.