Design And Realization Of The Analysis System Of The Natural Science Foundation Of The United States

In the era of rapid development of science and technology,the rate of integration and elimination is equally unprecedented.So seccurately grasping the research trends of related industries is an important requirement of research institutions and technology companies.The Natural Science Fund is not only the most intuitive embodiment of the official authority and social capital's judgment and evaluation of future technology,but also one of the most important sources of funding to promote the development of scientific research.So it occupies a very important position in the evaluation of the development trend of science and technology.As one of the largest natural science funds in the world,the Natural Science Foundation of the United States has an important position and significance.Fund analysis,that is,extract the important elements that users care about from the data of the fund,and describe the relationship between these elements and the process of developing the relationship.At present,the official website of the National Natural Science Foundation of the United States only has the ability to download and simple search for the past years of funds,unable to deeply explore the potential information of the industry and technology development contained in the fund.Fund analysis still requires a lot of manual reading and empirical judgment,which makes it unable to be carried out quickly,stably and accurately.In response to the above requirements,this paper designs and implements a fund analysis system based on the National Natural Science Foundation of the United States.The main workload is as below:(1)Based on named entity recognition technology,clustering technology and vector space model based on Bert,Achieved the goal of extracting hidden industry and technology keywords from the fund text.(2)Using an improved correlation analysis algorithm,establish a correlation graph structure based on the extracted fund industry,technology,related institutions,time and other dimensions,Show the relationship between important elements over time.(3)Designed and implemented an algorithm module combining offline and online,achieving the goal of optimizing the space occupation of the algorithm module and accelerating user response speed.(4)Realize data storage based on ElasticSearch and MySQL,Adopt Nginx-based front-end and back-end separation architecture,And use Kafka data queue to decouple the back-end data control layer,data processing layer and offline task layer,Realize the system's high availability,high scalability and flexible configuration.(5)Based on container technology and container orchestration technology,a one-click deployment and failure recovery mechanism of the system in a cluster environment is realized.This article comprehensively summarizes the design and development process of the system.We elaborated on the significance of the subject and the status quo of research at home and abroad in Chapter1.Then in Chapter 2 describes the technology related to this topic.Chapter 3 emphasizes the overall design process of the system and the selection of key technologies.We elaborate detailed description of the design and implementation of important modules of the system in Chapter 4.Finally in Chapter 5 we introduce the test process and results of the system according to the pre-set indicators,it proves that the functional and non-functional indicators of the system have reached the preset goals...