Folksonomy enables people to tag their information collaboratively with flat and free keywords. The outcome is that more people are participating into the process of categorizing knowledge without knowing or studying metadata schema like DC (Dublin Core)or DDC(Dewey Decimal Classification). The pros and cons of folksonomy are illustrated with the example applications such as del.icio.us,flickr and wikipedia. Finally, as a technology to organize information,folksonomy should adjust increasingly to the Web itself which is driven by people who use it daily.
The semantic environmental with special stop-words location information control has been studied and founded. This technology has been applied to Chinese metadata CXMARC text automatic indexing and the data mining of theme information. The algorithm of SWF that is used in the pretreatment special Chinese text automatic indexing can reduce the participle different meanings of a field efficiently and shorten indexing time. So tradition maximum matching algorithm has been improved of its quality and efficiency.
Web technology is continually developing with the appearance of Internet, Web, Grid etc. It can strengthen the ability to deal with digital resources. Therefore the services of digital libraries become better and better. Grid present itself and become the most advanced technology and the new trend of Web services development. Grid creates the possibility that digital libraries ot different areas could be integrated into global digital library for users. The article discusses service mode of digital library based on Grid, and brings forward the service characteristics and targets, the service modes and the realization of service etc.
This paper anatomizes some modules of Grammar Development Platform, which are characterized by complex algorithm. It introduces four functional modules, and concretely analyses algorithm of syntactic structure linear presentation, extraction algorithm of rule and lexicon, image creating algorithm of constituent structure tree and functional structure set, converting algorithm of constituent structure and functional structure. At last, it presents a simple test of grammar extraction subsystem.
As the first Ontology Server in the world, Ontolingua Server possesses good architecture and mechanism. This paper introduces the architecture, establishment and application of Ontolingua Server, as well as the mechanism of Ontology reuse. Besides, the advantages of Ontolingua Server are analyzed and compared with the current Ontology systems.
Along with the proposition of the concept of Ontology, the theory and practice of Ontology are gradually developing. The project of OntoWeb which is staked by the European Union Information Society Technologies has researched various knowledge of Ontology. Besides researching the correlative knowledge about Ontology, this project has also built a browsable prototype of the portal which is based on an Ontology. This paper introduces the architecture of this Ontology, its concepts and content types, the information provision, the browsing and the querying of the portal.
This paper chooses ten special Website of CDRS both at home and abroad, and analyses in detail the building of information resources and the development of operation of them , and then puts forward some ideas about how to construct the Website of CDRS.
This article introduces LOCKSS, a long-term preservation project of electric publication developed by the library of Stanford university. This article also discusses its system architecture, system functions, technologies and cooperation mechanism, etc.
This paper probes into the features, problems and improvements of the system which adopt Folksonomy as its information organization method through analyzing and comparing two Folksonomy-based digital management systems: del.icio.us and flickr.
As a large semantic knowledge resource, Ontology plays an important role in information processing. However, how to construct an effective Ontology is an important problem to its application. The first issue in automatic Ontology creation is domain concepts acquisition. In this article we experiment on a method to obtain domain concepts which are based on lexical cooccurrence (and then to support the automatic Ontology construction). The first step of this method is to obtain the primary starting concepts by manual analysis, and then to extract relative co-occurrence concepts from the corpus. Based on the corpus of People’s Daily, January 1998, the article focuses especially on the fields of sports and diplomacy. We extract the relative concepts, to examine the practical results of co-occurrence-based domain concepts acquisition.
We focus on how to make use of multimedia information more efficiently as the amount of such information increases rapidly. This article aims at speech, one kind of audio information. We combine Auto Speech Recognition (ASR) techniques with information visualization techniques, and then suggest a visual speech information retrieval model. At last, two speech/audio information retrieval systems are discussed.
Three kinds of centralized automatic discovery algorithms for network topology are introduced，which include building network topology based on ICMP or port scanning technology and utilizing SNMP protocol to visit in MIB（Management Information Base）to construct network topology. Then a kind of better network topology algorithm is proposed based on the three algorithms which can discover network topology accurately，unabridgedly and efficiently.
This paper analyzes the work theory and property of traditional searching engine in Web, the topological structure and property of P2P system and the structure and property of grid searching engine based on agent. lt also discusses the possibility of P2P system combined with grid system. The paper gives the improved model of grid resources searching engine based on P2P technique which proves to be resonable and has high efficiency by evaluation and analyse.
In this paper，we fristly survey and classify the current researches on personalized navigation of Web，and probe how to modeling the users’ navigation history by semistructured temporal graph, then point out its shortages. Also the paper presents an improved model which can great decrease the time and space complexity of the algorithm for realizing the model.
Virtual reference network is always one of the hot topics of the digital reference research. Different libraries are developing reference system appropriate to their own in China. The article outlines the new system of CSDL digital reference，such as: constitution of the system, service items, service mode and flow and so on, and discusses the present situation and the trend of the system.
Based on the construction of the evaluation index system of enterprises Websites, this paper uses FAHP to do the positive analysis of three tangible enterprises Websites, and evaluates the pros and cons of the application of FAHP in enterprises Websites.
The content of MODS is introduced firstly, and the reasons why we discuss the interoperability of MODS and MARC are given. Then the paper introduces the mapping of elements set between MODS and MARC, and the use of MARCXML as a middle layer of interoperation.
Combined with the basic datum of customer service operation of a domestic machinery manufacturing enterprise，an empirical study is given. Based on the SD conceptional model, the SD model of service system in the instance enterprise is constructed, and the validity of the model is tested. The simulative studies including information asymmetry control of product price, information asymmetry control of product quality, controls of service operation strategies and adjustment of information strategies are carried out, and the results are very important to the customer service operations of the enterprise.
This paper describes concepts, technology and policy chosen for data backup and disaster recovery. It lists fundamental, plan and policy of data backup and disaster recovery for digital library systems, and a successful experiment on disaster recovery based on open source software.
Based on a classed large-scale corpus labeled with key words, we can get words’ subject degree by statistic method according to domain erperts knowledge. The paper constructs a key words auto-indexing system, and introduces detailedly its overall process and functional module.
SciFinder Scholar is the biggest network database on chemical literature in the world provided by CAS. This paper introduces the general situation of the database, the retrieval method and characters.