This paper introduces the status of the digital reference book system in the university library and systematically explains the distributing problem of the teaching resource.Based on the problem mentioned above,the paper introduces a system,which can cluster and reconstruct the search result through the technology of the keyword index.

This paper discusses the semantic mapping expression and the similarity algorithm in knowledge grid integrated services based on Ontology,according to the particular function that knowledge grid describes the conception of knowledge and the relationship of them.Then an architecture of uniform integrated service platform with Ontology function is designed.

Based on investigation in literatures,technical standards and Internet documents,the paper analyzes the status of Ontology technology standardization,and then concludes that technical standards in top Ontology,domain Ontology construction,OMS design and repository are still insufficient.Finally,this paper points out that the Ontology technology should be standardlized.

This paper firstly builds medical domain Ontology on the basis of MeSH by Protégé 3.3 and KAON.Then the paper compares the quality and operation method of the two softwares from the basic information,building the basic structure of Ontology,visual management,storage management,and semantics Web supporting.

This paper introduces the basic idea of the Ontology-based IR,and puts forward a model of system according to the idea.An algorithm to enhance the relativity between the result and the aim is also given.

This paper analyzes the major applications of DOI,including the building of a citation-based open knowledge link system,the identification of digital objects in the distributed knowledge repositories under the open access environment,the link between science data and documentations, and content right management alone the industry value chain.Based on the above research,the paper also proposes the value and vision for promoting DOI in China.

This article focuses on the study of enhancing the function of major subject navigation system using Human-computer cooperation information filtering method.During the process of subject navigation construction,the method is used widely.Finally,a solution to the sustainable development of major subject navigation construction is given.

According to companies’practical demands on gathering and monitoring information on Internet,this paper firstly introduces the architecture of the Multi-Agent based distributed information collection system.A priority-based task scheduling algorithm is presented for central control agent in the cross-group level and for scheduling Agent in the in-group level.The author also proposes an in-group collaborative mechanism that can maintain load balancing among hosts where task Agents are executing.The test resuls for the system prototype demonstrate the validity and effectiveness of the priority-based load balancing task scheduling model.

TAM and TPB are the extensions of TRA. Due to the deficiency of these models being adopted alone,there are many studies on integrating them.A critical review to these researches is issued and a new model based on TAM,TRA and TPB is framed in this paper.

After analyzing the insufficiency of the existing E-commerce shopping patterns,this paper designs an intelligence shopping system based on Mobile Agent.It can help consumers efficiently and conveniently search commodity information which meets their requirements,recommend sound commodities and realize double win between clients and vendors.

It is indispensable to index stop-word before word segmentation in knowledge extraction.The key technique of processing stop-word is how to select stop-word,acquire and organize stop-word lists,and match stop-word.To recognize stop-word,constructing stop-word list is necessary.In processing stop-word,recognizing false stop-word can decrease noise.According to experiment,processing stop-word can not only save segment time,but also improve following syntactic analysis efficiency.

This paper analyzes the algorithm of Map/Reduce and uses open source Hadoop software to design high fault-tolerant,high-performance distributed search engines,which will be in the face of large-scale data processing and storage problems.

XFML,standing for eXchangeable Faceted Metadata Language, exchanges metadata in the form of faceted hierarchies. Its basic building blocks are topics,also called categories. XFML can’t solve all the metadata’s needs. It focuses on interchanging faceted classification and indexing data.The background,concept,and data representation method of XFML and principles of Web indexing with XFML are discussed in this paper. As XFML is a new concept to Chinese,using XFML will benefit the develop of Web information organization in China.

Statistics-based classification methods are common-used in hierarchical Web classification.However,classification precision of statistics-based methods often drops when categories are very similar to each other because of feature overlapping.Due to the nature of hierarchical Web classification,categories sharing the same parent (e.g.,leaf categories in the hierarchy) are often very similar to each other.To improve the classification precision,the paper proposes to use rule-based classification methods on top of statistics-based methods in hierarchical Web classification.Experiments show that our methods perform well on our education Web collections.

The enterprise-level search engine is proposed to solve the problem that enterprise have abundant document information resources but lack of effective search tools.The function and the overall framework of the enterprise-level search engine are analyzed firstly.Lucene indexer is studied in depth during implementation secondly. Then the plug-in unit is used to carry out the analysis and extraction of different types of documentsi in design. A set of parallel multi-task scheduling mechanism is established in the task scheduling. When the user interface is designed,Yui-ext components and DWR remote object invocation framework is applied to implement asynchronous communication by the Web,which can promote the users’ experience.

In view of the problems such as indexing slow,renewing no in time,searching inefficient in search engine of a unit’s Website,and based on analyzing and researching the merits of Lucene and XML technology on constructing search engine,this paper proposes an XML-based full text retrieval prototype system.The system uses XML as a universal data interface,uses Lucene as the development platform,thus achieves the purpose of rapidly timely indexing and improving retrieval efficiency.

This paper briefly analyzes the technology involved in whiteboard and introduces some knowledge about JGroups and data-share technology.Then the paper puts forward a new electronic whiteboard framework based on it,and finally presents the realization of the framework in practice.

In the field of large-scale information retrieval,distributed parallel IR technology is attracting more and more people’s attention due to its high efficiency and low cost as the result of the rapid development of high-speed network technology.Based on the design and implementation of BERSE,the paper discusses some key technologies for parallel IR systems,such as data distributing,query disassembling,and node sparing.

In order to solve the problems that education users have difficulty in getting resources and straightly link to the elementary education Websites, this paper mainly discusses the core technology of automatic finding and classification of elementary education resources and Websites.This paper also discusses heuristic rules and feather muli-selection method to find Deep Web.The prototype system using technologies mentioned above provides the education users with many classification ways for search.

Taking the application integration of the All-in-one Card System of Guangzhou University Library and Guangzhou University Town for example，the author introduces how to integrate the All-in-one card system without amending the library existing system.This paper briefly introduces the problems to be solved and gives a solution to them.The major technology used in implementation is given.