Knowledge organization and knowledge services are two international hard problems faced by digital libraries research. This paper discusses the background, status, role, basic theories of Knowledge Organization at Digital Libraries (DL-KO) and the research of Knowledge Organization System at Digital Libraries (DL-KOS) detailly, analyzing their current research deficiencies, and points out that Ontology-based and user-centered DL-KO is imprtant research in the coming day.

This paper discusses the knowledge base metadata standards on CALIS collaborative virtual reference service system. It introduces some basic issues about the metadata standards such as the descriptive units of metadata, the objects of description and their relationship, then introduces the structure, content and applications of knowledge base metadata standards. Finally it discusses related issues about knowledgebase metadata.

This paper introduces the background of SRU(Search and Retrieve via URL) protocol and why the OCLC SRW/U tool are employed in NSL(National Science Library of China). More specifically, the authors describe the architecture of the integration service platform based on SRU, the implementation of the SRU interface and the context-based service integration, the modification of legacy service systems in NSL. Finally, this paper introduces two applications based on the SRU server in NSL: the user desktool and the new Website of NSL. With the same techniques described here, other libraries can create their own SRU server and provide the embedded services for their specific needs and audiences.

This paper proposes a retrieval model of digital libraries based on the ChinaGrid Support Platform(CGSP). The authors try to utilize the CGSP as a grid middleware to integrate existing information resources, hardware and other software as well. In the end, this paper discusses relevant key technologies including grid services and typical job.

Collaborative filtering recommendation systems in digital library have faced the problem of sparse user ratings. To solve the problem, a computing method of group interest trend degree has been proposed and used into the prediction of vacant values in user-item matrix. The experimental results show that the algorithm can efficiently improve recommendation quality.

This paper demonstrates the relationship between adaptability and life cycle of system, and analyses the key factors among system architecture attributes which have great effect upon adaptability of system. At last, the authors construct a prototype of information service based on SOA.

This paper presents a research content framework for Web digital information seeking, and extends the framework based on a through review of the 3 research branches,Web information retrieval,Web information seeking and online consumer information seeking.Several problems for future research are also identified through.

According to the different managers of Web Services,this paper proposes three possible operation mechanisms of Web Services:Web Services predominant characteristic mechanism gudied by large enterprises,Web Services portal mechanism guided by industrial category and Web Services Uniform Alliance(WSUA)mechanism.On the basis of comparison of these three mechanisms,the authors think that the ideal operation mechanism is building a Web Services Uniform Alliance.In the end,this paper gives the methods for getting Web Services reasonably and legally.

Based on the studies of system architecture of NLP platform and knowledge extraction system, the author brings forth a detailed resolution on how to design a knowledge extraction system based on NLP. NLP technique includes eight modules, such as segmentation, part-of speech tag, syntactic analysis and semantic analysis. Knowledge extraction includes four modules, such as documents type analysis, discourse analysis, knowledge extraction and knowledge representation. Research on system architecture of knowledge extraction based on NLP is beneficial to not only find relations between NLP and knowledge extraction, but also analyze system flow and critical technology of knowledge extraction.

Feedback noise in image semantic retrieval is a key issue in this paper. Firstly, the bad effect of feedback noise on the semantic network method is analysed. After that, an algorithm of relevance feedback based on voting idea is proposed, which is robust to feedback noise. Then an analysis is made on the performance of the proposed algorithm. Finally, further study issues are pointed out.

In order to better apply association rule mining technique to query expansion and find out some better query expansion models, 4 categories of query expansion models with 13 varieties are given based on item-all-weighted association rule mining. Comparison of retrieval performances are made through experiments. Some better query expansion models are discovered.

Through multiple linear regression analysis of the data collection from user satisfaction survey for electronic resources, the relationship between user satisfaction and impact factors is modeled. The regression equation between overall satisfaction and 15 impact factors is established by significance testing, correlation coefficient analysis, and multicollinearity diagnosis. A strategic matrix is worked out.

This paper uses association rule mining algorithm to analyze the database and to recommend related document. Meanwhile, this paper proposes mixed weighted association rules that fit to recommend related document. This paper identifies the related document and the vertical weight by analyzing users’ behavior.The authors use the Google’s PageRank algorithm to define the documents’ horizontal weight and obtain some meaningful results.

In order to meet the demand on the data-mining under the E-commerce environment,this paper copes with concrete problem with Clementine,and puts forward the streams through data-mining,including the understanding of apply and data,as well as preparing and models’ application. Clementine has enormous development prospects to the discovery of information under the E-commerce environment.

This paper discusses the difference between XML retrieval and traditional information retrieval, the aims and tasks of XML retrieval and the key problems of the research on XML retrieval system. Then the paper introduces and carries on a comparative research on some typical XML retrieval system.

Based on real-time metadata management,a software architecture for the LIMS is proposed in this paper.The architecture integrates heterogenrous experimental system,using expanded Web services mechanism.Using model management operators,schema evolution is discussed in the high abstract level in implementation.Web services composition is shown by a business process example.It meets the requirements of integration of an economics and management laboratory.

This paper analyses the problems existing in the semantic matching methods in the typical distributed UDDI networks. By expanding the classical elastic matching algorithm,the author uses the GCSM semantic distance algorithm, class factor and category factor to present an improved matching algorithm of semantic Web services.This algorithm can be used to compute the match degree of Web services so that the match results approach the request.

The basic theory and its features about Latent Semantic Indexing(LSI) are analyzed.For the three factors of LSI, the word selection,dimension simplification, words weighting have been engaged and improved. Scientific and technical literatures from computing are used as testing documents, also the improved weight algorithm and the retrieval results about two LSI systems are analyzed. The experimental results show that the feature choice and retrieval results are superior improved and hard performance with the new weight algorithm.

This paper compares some usual methods of processing MARC in ISO2709,and gives a new data structure of nested hashtable and dynamic array to normalize MARC.The author shows how to sort the MARC data according to this data structure and delete empty fields.