This article mainly discusses the theoretical methods and technologies of information visualization prototype system. It specifically discusses constructing strategy of visualization model, environmental configuration, functional module and operation method of prototype system. What’s more, it also discusses construction of text, voice (audio) and image information visualization model, data preparation and data scale, operation interface and running result. The research on this prototype system is not only a helpful attempt in studying visualization model of general information resources management with certain experience achieved, but also a successful explore in terms of visualization of Chinese information.
Based on situation awareness theory, this paper discusses the relationships between situation awareness and visualization, then a visualization awareness model is constructed. The model consists of 5 stages, including situation awareness requirements analysis, extraction of data and knowledge, situation visualization & interaction, situation awareness, and decision making & enforcement. At last, the key problems of visualization awareness are presented.
According to the actual needs, the article puts forward a new method which applies moderately simplified Chernoff faces on multi-dimensional strategic data analysis. The strategy in the index variables distribution of Chernoff faces is the most important index distribution method based on entropy weight. Using improved Chernoff faces as visualization tool, the authors make a practical comparison analysis on some key economic strategic data in 2008 of three prefecture-level cities within the Pearl River Delta area. These cities are Huizhou, Zhongshan and Jiangmen. The research result shows that the method has the validity and practicality.
This paper dedicates to visualization mining methods of topic data model under eTOM(enhanced Telecom Operations Map) framework with some telecom company as the research object. It explores new theories and implementation methods in enterprises informationization. In the implementation process, business process analysis and visualization mining are achieved through Pajek and TreeMap. Detailed visualization mining implementation is discussed in human resources topic data model.
This paper introduces the concepts of information security in digital library and summarizes the findings of academic circles at home and abroad. Based on the survey of 30 public libraries and college libraries which have established digital libraries, this paper analyzes and summarizes the current situations and problems of information security management in domestic digital libraries.
In accordance with the ISO27000 series standards, this paper proposes the measures of digital library information security risk assessment. On the basis of distinguishing and evaluating the digital library assets, threats and vulnerabilities, this paper calculates and analyzes the risks which digital library faces with, puts forward the measures of dividing risk levels, and forms the model of digital library information security risk assessment.
According to characteristics and the realistic need of digital library, the paper takes the common criteria of ISO27002 as a standard, summarizes some control measures for information security management of digital library. Then it screens out the core control elements and referenced control elements about information security in digital library, and carries out risk control on digital library from organization control and technology control.
Based on a full investigation of user-led, user-driven and user-centred digital library，the article introduces details of user model that guides the information system functions, processes and interaction. The recommendation is a concise method as “user psychology and behavior - digital service- information system - market research - improve information systems” in order to optimiz digital library service system.
Based on the analysis of recent related literatures and projects, the paper concludes the term semantic measure methods as follows: similarity measure methods based on corpus characters and similarity measure methods based on open knowledge resources. And then it reviews the integration methods of multi-measure methods. It also summarizes the applications of term semantic similarity measure methods on the area of Natural Language Process (NLP) and Knowledge Mining (KM). Finally, the future development of research on term similarity measure is discussed to help build more efficient term similarity calculation system.
This paper makes a research on the implementation of network-based retrieval and reasoning about Ultra-Large-Scale OntoThesaurus, and the proposed solution has successfully applied to the realization of the CCT1_OTCSS, which is a co-construction and sharing system of an Ultra-Large-Scale Ontology named CCT1_OntoThesaurus. This paper proposes the structure of Lucene index based on the idea of triple “subject, predicate, object” of the RDF, and validates the feasibility of implementing efficient retrieval, terminology services and reasoning based on the Lucene index of Ultra-Large-Scale OntoThesaurus. The solution can be reused for several Ultra-Large-Scale Chinese thesauri most widely used in China at present, implementing quickly Ontology-oriented upgrading, networked co-construction, sharing and dynamic updating for them, and also has a reference value for other large-scale knowledge organization systems (thesauri, Ontology, etc.) in the form of XML, RDF or OWL at home and abroad.
To cope with the problems in mainly manual construction of Knowledge Organization Systems（KOS）, such as shortage of professionals, tedious work and difficulty to guarantee consistency etc., automatic enrichment of KOS based on scientific and technical definitions is proposed. In this paper, key techniques of definition extraction, definition resolution and definition-knowledge conversion are studied. It is proved that the relative studies have a great value that they can support and affect the construction and updating of scientific and technical KOS in the future.
This paper compares the advantages and disadvantages of the knowledge representation methods, and then introduces the superiority of Ontology based knowledge representation and knowledge reasoning emphatically. At last, it illustrates the applicational advantage of Ontology in expert system by constructing history domain Ontology model.
Based on Conditional Random Fields（CRF）, the article identifies the Chinese prepositional phrase. In order to identify the prepositional phrase effectively, the article counts and analyzes the external and internal linguistic features of the prepositional phrase. The prepositional phrases without nesting and nested prepositional phrases are identified with the complex feature model,and in open tests,the best F value can reach 90.29% and 89.99% respectively.
This paper first introduces the current status of analyzing core technology, core applicant or inventor in patent analysis system. Then it analyzes the feasibility of applying the Bradford’s Law to determine the core technology, core applicant or inventor. Lastly, it uses two specific examples for verification.
According to the deficiencies of K-means clustering，the paper proposes a modified clustering algorithm which uses SD and PSO algorithm, and achieves this integrated algorithm in Java. In the analysis, the authors take customer transaction data of a securities company in Shanghai as an example. By transforming the database into a form suitable for mining, the paper applies the modified clustering algorithm to cluster segmentation model, and the clustering results show that the improved clustering algorithm can get higher quality clustering results.
This paper proposes an anti-shielding solution integrated with different technologies to avoid shielding, improves Web content extraction based on text density, adopts eliminating duplication technology based on VSM and cosine angle formula, and develops a system of the Internet subject acquisition system on overseas Chinese.
This paper firstly designs the framework and process of local digital resource system based on open source. Then related open source technologies and services are analyzed. The system designs the unified local processing mode in various kinds of resources, and realizes unified search service of local resource based on Lucene and unified search integration services to portal system based on Web service.
This paper proposes a new method for calculating similarity between secondary documents of National Science and Technology Library（NSTL）, subsequently develops and implements a subsystem of real-time related documents recommendation. At last, a measurement which can theoretically evaluate the quality of recommendation results is presented. And the experiment results demonstrate that the system can improve the service quality of document retrieval system of NSTL.
For ubiquitous subject indexing problem of Chinese books in small and medium-sized libraries, this paper uses the FP-growth algorithm to mine discipline subject field, and the discipline thesaurus of each category is obtained. The reference basis for making machine-readable data subject indexing rules and bylaws of Chinese books in small and medium-sized libraries is provided as well as subject indexing of Chinese books.
Based on the TRS full text retrieval platform, this paper presents a Three Famous Indexes Paper Data Management System designed by B/S architecture, TRS Java API and open plug-in technology such as JFreeChart and POI. This system realizes a Web-based data management, which provides such functions like information retrieval, data maintenance, graphical statistic and Excel generation.
This paper improves the composing of library digital reading room management system based on the all-in-one campus card. The authors research on the characteristics and deficiencies of charge pattern, and make some improvement. The improved charge pattern can solve the problem of cost evasion when the users logout their all-in-one campus card.
