A Study On Scientific Data Management In University

This object of the study is scientific data management in university. The Grounded Theory is adopted in this research. Based on the background analysis and a large number of documents study analysis, combined with the library science and information resources management related theory and academic information resources management practice, this paper discusses the principle of scientific data management in university, constructs the mode, designs the technological route, and carry on demonstration study.Firstly, a comprehensive collection from library and information science, and also from the scientific community, which involve in all aspects of the literature material were reviewed and discussed, from which to obtain scientific data organization related theory, method and case, and in accordance the concept of scientific data under the background of e-Science.Then, on the one hand, from macroscopic perspective management system as the external environment, combined with the library science and information resources management theory, the life cycle of information organization theory, information description theory and information select theory, as well as the theory in the micro level of the process of application, discusses the principle of scientific data management in university; On the other hand, through the analysis of research activities and scientific data characteristics, and the role of academic library, build the scientific data management mode based on university institutional repository.In this mode, the implementation process, through the analysis of the different characteristics of the scientific literature and scientific data obtained from the analysis of the specific case of economics and materials science, it is concluded that the scientific data as information life cycle and mapping of scientific research activities in the life cycle of its two characteristics, in accordance with the RDA (Resource Description and Retrieval) standard design scientific data resource description framework, and details of the metadata scheme, and try to implement.This paper studies in the following section.The first chapter, defines of the research background and the topic selection, summarizes and comments literature from the scientific community and the LIS, introduces the object of study, the study route and research methods.The second chapter, analyzes the university scientific data management environment, first from scientific research object changes, scientific research complexity and information technology development perspective of the origin of e-Science, the concept, technology, characteristic, significance and relevant factors for e-Science, briefly describes Britain, the United States, Europe and China's practice. Then discusses the emerging of data science, the "data intensive" scientific research paradigm under the background of scientific data definition, and many problems involves e-Science.The third chapter of the study is a scientific data management principle. In this paper, the principle of scientific data management is divided into macro and micro levels. The macro-level does not involve the details and process, to consider only the external factors related to the management of scientific data, scientific data management system, this paper considers the information policy of the national level is the top-level environment for scientific data management, scientific data management is intended to establish sharing system, including standards to develop the work of the national or international specific field of expertise or specialized agencies, the development of key technologies, platform architecture; publish as an effective form of professional management and sustainable development of scientific data and means the final part of the university scientific data management; and data scientist as human resources for scientific data management. Micro-level includes scientific data management process, all aspects of the research management process, the various aspects of the relationship. The scientific data will become the main part of the future data digital library, the effective operation of the digital library, information organization and description is the key. In this paper, within the scope of scientific data management is mainly aimed at scientific data characteristics in implementing the information organization, and finally provides sharing access. Scientific data as academic information resources, and mapping of research activities of the two layers of the life cycle meaning, this is the description of the basis of scientific data. Scientific data archiving and distribution is an important part of scientific data management. What scientific data need to be archived needs the support of the theory of information selected.The fourth chapter is the content of the university scientific data management mode. University research activities have the characteristics of "contemporary small science" and combine the features of teaching and research. To form the results of research group as the "root" of the "research achievements tree", which scientific data as part of it. Also because of the characteristics, the scientific data presented features of individual research projects and cycle development, circular discipline class and its scientific data type and format diversity, science data management of "non-planning", need special organizations in campus unified management. University library have traditional intermediary in the academic information resources transfer, and the platform of academic communication function also shows more and, at the same time, the library in the long-term of the academic information resources management has accumulated rich experience. Further, the institutional repository maintenance by library, which preserve university research output to be published and not fit for publication. So, it determine that the institutional repository of the university as the scientific data management mode.The fifth chapter is the content of the technical route. Through analysis of the correlation of retrieval behavior, scientific research activities and information organization, this paper concludes that the user research activities orientation management mode, which combined the information resources organization with scientific workflow. Life cycle management of scientific data as information resources, but the more important is life cycle as mapping research process which is information continuum management. The meaning of this model will help researcher understand the scientific process when they retrieval and reuse the scientific data. Because of these characteristics, and the distinction between scientific data and literature, this paper using RDA as standardize for the description of scientific data, which the theoretical basis are "Functional Requirements for Bibliographic Records"(FRBR) and "Functional Requirements for Authority Data"(FRAD), and FRBR and FRAD were supported by E-R conceptual model. Through description of the entity which includes scientific data and scientific data sets, and description of the relationship of entity, the composite object of university scientific data management possible.The sixth chapter is the content of empirical research, which is also the implementation of technical route. Base on the case study of research activities with scientific data from economics science as social science research and materials science as engineering research, comparative analysis research activities, and this paper verifies the characteristics of research activities and scientific data. DSpace system is selected as institutional repository platform. Through analysis of the system structure and information organization model of DSpace, according to the characteristics of university science data management and requirements, this paper design the scientific data information organization solution of "research group-scientific data set-science data" corresponding to "community-collection-item", which is DSpace system information organization model. According to the technical route, the demand of description scientific data (set) and the mutual relationship, the quality DC metadata, the paper design the metadata scheme for scientific data (set). And write example accordance with the "Metadata Encoding and Transmission Standard"(METS), which is the encapsulation of the management metadata, structure metadata, and technology metadata.This paper study university scientific data management environment, university scientific data features; university scientific data management features, and then draws a conclusion. In the end, the conclusion has been verified through examples.The innovation of this study is mainly reflected in two aspects:(1) two layers of life cycle management for scientific data management, the first one is scientific data as information source, the second one is scientific data as the mapping of research activities;(2) the information description particle degree is scientific data records not scientific data.