Construction And Application Of Agricultural Scientific Data Curation Model

In recent years,China's agricultural scientific research has made rapid progress,generating a large amount of valuable agricultural scientific data.These scientific data relate to all areas of agricultural science,and researchers and agricultural research work have a wide range of needs.However,because agricultural science data is mostly not network data,it cannot be obtained through the “natural formation” of the Internet,but “creation” in scientific research work.It is hard to come by.Many scientific data require specialized observations,experiments,and excavation of professionals and equipment.The investment is large and time-consuming,which creates obstacles for agricultural researchers to obtain agricultural scientific data.Affect the effective use of agricultural science data.The rapid growth in the amount of data resources in agricultural science has brought both more opportunities and greater challenges for researchers to obtain the information and knowledge they need.Due to the diversification of agricultural scientific data sources and expressions,it is difficult for agricultural scientific data to have a standardized storage format to ensure the integrity of agricultural scientific data.Agricultural scientific data regulation is not simply the storage of agricultural scientific data,but the ongoing curation of agricultural scientific data for the life cycle of academic,scientific and educational purposes.By evaluating,screening,reproducing and organizing data for current agricultural research activities,and for future rediscovery and reuse,it can construct effective scientific data resources for solving decision problems in agriculture.It provides new ideas,new methods and new ways to solve the problem of agricultural scientific data curation services in the field of agricultural scientific data resources.This paper aims to achieve agricultural scientific data reuse and data appreciation as a foothold to meet the scientific research of high-quality scientific data curation services.Comprehensively use theoretical methods such as information science,agricultural science,data science,and computer science,and propose to structure the agricultural science dataregulatory framework model from the perspective of scientific data organization.Construct a data acquisition process model and propose a theoretical framework and solution for centralized curation of multiple source,decentralized,and submerged agricultural science data resources.In the big data environment,this paper attempts to explore the implementation methods of agricultural scientific data curation from three different dimensions: scientific data curation influence factor dimension,life cycle dimension and service dimension.Specific research content includes:(1)Through the elaboration of related concepts such as agricultural scientific data curation,the scope of research objects and research contents is further clarified.At the same time,this paper traces the source of digital agriculture theory,life cycle theory,data science theory,ontology theory,machine learning theory and knowledge discovery theory which have important guiding significance and reference value.Among them,digital agriculture theory,life cycle theory and data science theory provide important theoretical support for the construction of agricultural scientific data curation model.Ontology is a powerful tool to realize data organization in agricultural scientific data curation,while machine learning and knowledge discovery theory It provides technical solutions to solve problems such as classification and aggregation of agricultural scientific data.(2)The objectives and principles for the curation of agricultural scientific data resources were determined.The scientific data needs of users(scientists)and the data scientists(data regulators)were analyzed to meet the data requirements of agricultural scientific data,and the regulation of agricultural scientific data was clarified.The basis for the framework of processes and architecture.Then it analyzes the relationship between the constituent elements,functional elements and elements involved in the agricultural scientific data curation process,and proposes the logical framework of agricultural scientific data curation.(3)In the dimension of influencing factors,qualitative research methods were used,and the Grounded theory was used to analyze the influencing factors of agricultural scientific data curation.Through the model of in-depth interviews,the interview outline was designed.Using Nvivo software to record,organize and analyze the information of each interview,analyze the data through coding form(open coding,spindle coding and selective coding),form a model of agricultural scientific data curation influence factors,and explain the model.(4)In the life cycle dimension,from the perspective of the life cycle of agricultural scientific data curation,firstly,through the data collection and analysis of agricultural science,taking the deep learning which is included in machine learning as an example,the complexity of agricultural scientific data collection is pointed out;Secondly,it analyzes the organization of agricultural scientific data,organizes agricultural scientific data by ontology method,uses Protégé software in ontology,and realizes the effective organization of agricultural scientific data through agricultural scientific data meta data model.Finally,through the agricultural scientific data sharing model,the ultimate goal of agricultural scientific data curation is achieved,and the agricultural scientific data is maximized and reused.(5)In the service dimension,the paper discusses the construction of the agricultural scientific data curation service model.Based on the stakeholder theory,through the statistical analysis of the agricultural scientific data curation stakeholders,the three core stakeholders of government,users and data service personnel are identified.Based on these three dimensions,we find the core concepts of the three main models of government policy,user demand and service model.At the same time,we analyze each dimension in depth,and through the relationship analysis between these three dimensions,we have constructed The user-oriented agricultural scientific data collaborative curation service model is expected to guide the specific service practice of agricultural scientific data curation through the establishment of this model.(6)According to the structure and function division of the multi-dimensional framework of agricultural science data curation,the methods needed for the realization of agricultural science data curation in different dimensions are proposed,and the effects of these methods need to be verified through application.In this paper,the sky to the integration of agricultural curation system of data resources as an example,analyzes the data in different stages of the life cycle(data collection,data organization,data storage,data sharing),the application of the "sky integrated curation system for digital agriculture" in thestructure are analyzed,and put forward the tactics of improving the quality of agricultural science data curation.