Research And Implementation Of Knowledge Graph And Recommended Method Based On Qinghai Science And Technology Big Data

With the rapid development of science and technology,a large number of service demands have emerged in the fields of scientific and technological innovation and the application of achievements,and the role of science and technology service departments in the modern service industry is becoming more and more important.In the case of a large number of scientific and technological information resources,how to recommend the scientific and technological service resources they need for scientific and technological managers and scientific researchers has become an important issue.The existing scientific and technological service system usually adopts the traditional search recommendation method based on collaborative filtering or content-based,lacks the utilization of knowledge fusion and knowledge association,and has the problem of poor interpretation and cold start of traditional recommendation methods.Knowledge graph is a new type of knowledge base system built by combining big data and knowledge engineering,which is easy to realize the fusion,semantic retrieval and inference recommendation of multi-source heterogeneous data,and can provide a technical foundation for scientific and technological big data analysis and intelligent recommendation.Based on qinghai science and technology big data,this paper carries out the design and construction of scientific and technological knowledge map,and provides method support for accurate recommendation of scientific and technological services.The main tasks of this article are:(1)According to the Qinghai Science and Technology Big Data Platform,the qinghai Science and Technology Big Data Knowledge Graph was built,the conceptual model of the knowledge graph was constructed,the specific algorithm was written in Python language to complete the knowledge extraction work,the support vector machine(SVM),Naive Bayes(NB)and K nearest three machine learning algorithms were used to train the classifier model,the entity alignment experiment was carried out,and the naive Bayes algorithm with good effect was selected through the comparative experiment to complete the entity alignment work of the researchers.Finally,the graph database Neo4 j is used to complete the knowledge store work.(2)The Ripple Net recommendation algorithm in the field of e-commerce is applied to the field of science and technology big data recommendation,combined with the characteristics of scientific and technological big data,the Ripple Net algorithm is improved,and the Adjust-Ripple Net algorithm is proposed,which replaces the user's historical behavior data in the Ripple Net algorithm with the user's background data,constructs the user-background interaction matrix,and classifies and recommends the recommendation results according to the tags of the entity nodes in the knowledge graph.And the design comparison experiment confirms the advantages of the algorithm.(3)Combined with the construction of Qinghai Science and Technology Big Data Knowledge Graph and Adjust-Ripple Net recommendation algorithm,the Qinghai Science and Technology Big Data Support Service and Decision-making Platform is designed and implemented,which contains three modules: user login module,knowledge retrieval module,and knowledge recommendation module,the system adopts B/S architecture,the functional logic is deployed on the server side,and the user can enter the system through the browser to obtain services.