Research On Semantization Of Digital Literature Resource Based On Metrics Analysis

With constantly deepening of economic globalization and the rapid development of science and technology, our world is entering a whole new era of knowledge economy. Knowledge has become important strategic resource replacing the traditional factors, such as capital, labor and so on. Knowledge innovation is an important part of the national innovation system, and the key to obtain a competitive advantage; knowledge communication is the prerequisite and basis of the knowledge innovation and an accelerator to promote the development of science and technology and the knowledge economy; knowledge discovery is the fundamental and critical source of knowledge innovation; knowledge service based on library resources is the fundamental support to jointly achieve knowledge innovation, knowledge communication and knowledge discovery in scientific research. However, in the social context of the large-scale development of the knowledge economy, science and technology progress, cultural industries prosperity, there are many issues that must be in-depth research and comprehensive solution, in order to deepen knowledge services and continuously improve the capacity and efficiency of knowledge services and information services in libraries.Facing the national innovation system and the development of the knowledge economy, and according to the basic principles of the organization of knowledge from the digital library, this study using the integrated methodologies and tools related to ontology and semantic, informetrics, network science and other fields, innovatively proposes a semantization theory and application model of digital literature resources based on informetrics analysis, and tentatively constructs the metrics-ontology model for knowledge organization and retrieval application in digital literature resources; then makes empirical analysis to verify the operability and practicality of the semantic ontology model, and fully realizes the semantic association revelation and knowledge reorganization and polymerization in digital literature resources based on the metric relationship. This research seeks to improve the efficiency of knowledge organization and knowledge sharing in digital collection resources based on the metrics-semantic ideas, aiming to improve capacity and utilization of library knowledge service, in order to provide a useful guide and help for the future development of digital libraries and construction of interdisciplinary methodology system.Adhering to the principle of "combining theoretical research and practicical application", this paper conducts the digital literature resource metrics-semantization research from the aspects, such as theory exploration, semantic model construction, metrics-ontology putting forward, empirical research, and so on. Excluding the introduction and conclusion, the paper consists of five chapters and each chapter is self-contained. The specific contents are as follows:The first chapter aims to comb and discuss the theoretical basis related to digital literature resource semantization. Firstly, the basic theories of semantics, ontology and metadata are elaborated, such as the concepts, principles, applications and their relationships. Secondly, the research object, i.e., digital literature resource, is defined; the connotation of digital literature resources organizations are analysed; organizational models of digital literature resources are summarized. Subsequently, a detailed analysis about the relationship between semantization and knowledge organization and the influences for digital literature resources organization from semantic technology is carried out, and the basic ways of digital literature resources semantization following the two semantic formats of content and organization are expounded. Finally, the issue about digital literature metrics-semantization is put forword in view of the presenting problems and the trends in the current resource semantics.The second chapter mainly dissects the theories of digital literature resources semantization based on metrics analysis. In the first part, an analogy analysis for metrics analysis and semantic ontology is made in three levels of theories, techniques and methods as well as the application. We find that they can correspond to each other in composition principle, expression form and theoretical connotations, and there are also some similarities in their methodology system, technical principles and analysis path, and they also have some in common in the aspects of implementation process, application ways, results show and so on. These common characters of metrics analysis and semantic ontology produce the possibility of introducing the metrics analysis into digital literature resources semantization. In addition, in view of complementary advantages of improvement for semantization efficiency and integration for interdisciplinary approaches, it's also very necessary to introduce metrics analysis into resources semantization. Therefore, this chapter presents a theoretical framework for digital literature resources semantization based on metrics analysis, which provides theoretical support for building the metrics-semantization application model.The third chapter reveals the mechanism and mode of digital literature resources semantization based on metrics analysis and constructs a metrics-semantization model for digital literature resources organization. In the first part, the generation and diffusion of metric-semantic relations in digital the literature resources are discussed in accordance with the integrated principles of metrics analysis and semantic ontology; metric-semantic association calculation path is conceived with the combined use of informetrics relation analysis, vector space model, and Bayesian network method; thus, a metric-semantic network consisting of concept nodes, relation arcs and corresponding identifiers is further resulted; at the same time,11core reasoning rules of the metric-semantic network are proposed blending in the deduction mechanism of metrics relationships. In the second part, the five modes of digital literature resources metrics-semantization are refined, i.e., basic semantic mode, multi-layer semantic mode, multiple semantic mode, the high-level semantic mode, multidimensional semantic mode. Finally, three application ways of digital literature resources metrics-semantization are set forth, i.e., digital literature collection integration, semantic information retrieval and recommendation, along with knowledge organization and knowledge services; and the metrics-semantization model for digital literature resources is constructed, which is composed of five modules:digital literature resources metadata module, informetrics and statistical analysis module, metric-semantic analysis module, metric-semantic knowledge extraction and discovery modules, metrics-semantization application module.Chapter IV presents the concept of metrics-ontology for digital literature resources and defines the basic building path for this special kind of ontology. Firstly, the comparisons and discriminations among this innovative concept of metrics-ontology and traditional domain ontology, and the ontologies in broad sense or narrow sense are respectively carried out, and the similarities and differences between metrics-ontology and traditional ontology are pointed out. Subsequently, the basic elements of digital literature resources metrics-ontology are illustrated, including classes and instances, attributes and values, constraints and inference rules; and the overall framework of metrics-ontology composed by abstract semantic concept layer and metric-semantic association reasoning layer is put forward. Finally, the eight-step process of metrics-ontology construction is summed; and the core architecture of metrics-ontology is built and initially expanded by Protege tool; at the same time, the storage mode of metrics-ontology is also presented, whose core elements are metadata tale, class table, the attribute table and attribute-instance table.Chapter V focuses on the empirical research of digital literature resources semantization based on metrics analysis. There are three aspects of semantic empirical application to be explored respectively in plant science field, information science field and scientometrics field from publication angle, citation angle and integrated angle accordingly:the first one is semantic description application including metrics-association attributes reasoning, association degree measurement and ontology construction; the second one is semantization and visualization analysis and the application of associated structures deduction and discovery, which are based on metrics analysis, relation extension and reasoning, association measurement, factor analysis and ontology construction; the third one is a combining application of semantic network analysis and presentation, association mining, semantic retrieval and recommendation, according to the fully associative reasoning, social network analysis, semantic reasoning and mining. These research results can provide new ideas for the discoveries of research themes and peers in the corresponding fields, which can also further extend the new paths of the future knowledge services in digital libraries.