Based on the application practice of sustainable knowledge-based subject groups integration platform carried out by National Science Library of Chinese Academy of Sciences during the National Twelfth Five-year Plan period, the paper discusses strategic infrastructure and construction content for the sustainable development of this platform in the views of development dimension, collaboration capacity and sustainable magnititude.The integrated and comprehensive solution of the sustainable development of the platform is proposed from the perspectives of information resources, information services, content-based organization tools, and information-sharing tools.

This paper puts forward a solution of the institution group integration knowledge platform based on principal and subordinate model after analyzing the open source project Liferay. Several independent group platforms are harmoniously integrated into an institutional knowledge service network using such technologies as information extraction,customized push,RSS aggregation and interface acquisition. Finally, the main method of secondary development is introduced.

Subject Knowledge Environment(SKE) is a tool to build the knowledge environments and academic communities embedded in the user scientific workflows. Based on a research Ontology and a domain Ontology as the knowledge organization models, and the application on semantic Web technologies and reasoning rules, SKE has the functionalities on information management, knowledge organization and knowledge discovery, while realizes the user customizations on Ontology reusing, system management and data reusing. SKE is currently in the services on building the XKEs, mainly used in the constructions of specific subject knowledge environment, research community, project information environment and collaborative research platform of major research groups.

The dispersion of scientific resources leads to the issues that the researchers obtain the information singularly and use the open resources complicatedly. To solve these problems, the method of designing ISCAS group platform based on resource integration, knowledge organization and service integration is proposed, and the practice that keeps the platform developing sustainably by using supporting tools is presented.

Based on the Subject Knowledge Environment(SKE),the paper tries to construct scientist network in plant sciences research with the requirement of knowledge discovery of Shanghai Chenshan Plant Science Research Center.Through researching and developing key technologies including Ontologies, user interface and video filming,a plant scicence knowledge enviornment with personalized characters is built and applied.

This paper reviews the 13th International Conference of Asia-Pacific Digital Libraries(ICADL 2011), introduces the main contents and trends of the following research areas, including digital preservation and management, data mining and application, mobile services, metadata and catalogue, parallel computing /cloud computing, social network and personalized service, and information retrieval.These contents can represent the main trends of Asia-Pacific digital library research and application.

This article analyzes the basic principles and technology stack of linked data. It provides an overview of implementation about linked data service of OTCSS based on Chinese Classified Thesaurus, and introduces the service functions of linked data in detail, further makes comparison with linked data realized based on LCSH.

To avoid waste of resources during construction of digital library, Xen Cloud Platform is suggested as infrastructure virtualization platform. Xen Web manager is selected to manage the resources effectively in Xen Cloud platform. Some value-added services developed with open source framework such as operation management, resource management, security management and Xen extension are designed to solve the key problems in the process of library private cloud construction. Thus an open source private cloud infrastructure solution for library is laid.

Considering the ignorance of readers’ information in the book retrieval of library, the paper proposes the hybrid retrieval ideas based on hybrid thinking and creates a new retrieval model. According to the retrieval frequency of the books, the new retrieval model divides the books into “cold or hot” categories, which turns the one-way retrieval into two-way retrieval. Combined with the characteristics of the readers and the use of Apriori and K-means algorithm, it builds a mathematical model and provides readers with the most appropriate books. In accordance with the new ideas, experimental data is used in the comparative evaluation of new search and the old search,and the experiments show that the retrieval accuracy and efficiency of the new models are improved significantly.

On the basis of analysis of modes and standards to access terminology, this paper puts forward a system structure of the terminology registries and terminology services based on REST, and constructs an actual platform for terminology registries and terminology services. This platform provides such functions as terminology converting and storing in batch, terminology registering and accessing, terminology visualizing and Web services API for users. And then this paper discusses the characteristics of the system structure and the advantages of this platform. Finally,it summarizes the development trend of this field.

In allusion to the limitations of current researches of international academic community for UBKD in Folksonomy, a FCA-based UBKD model in Folksonomy is proposed in this paper. The model consists of seven functional modules:problem definition, data acquisition, data preparation, data organization, data mining, knowledge generation and evaluation & feedback. The core idea of the model lists as follows:on the basis of data organization which using the “F^U:= (U,T×R, Y^U)” context, a visual data mining of common knowledge and personalized knowledge for UBKD in Folksonomy based on concept lattice is respectively achieved by the method called “external mapping” and “backtrack”, and finally the user group hierarchy and the single user behavior chain are respectively constructed.

This paper proposes a dynamic faceted method based on formal concept analysis, which extracts author keywords of literature, then merges the semantic similar keywords based on similarity calculation, and generates different granularity formal contexts. Finally, a lattice-based semantic faceted navigation is given. The empirical analysis shows that the method is feasible and effective.

This paper provides a comprehensive review on word similarity measuring methods in two categories, with background knowledge and without background knowledge. The statistical method without background knowledge cannot reveal the semantic relations between words, and the thesaurus provides limited scope of words. Wikipedia, as a large corpus comprising semantic knowledgebase, becomes the new sources for measuring semantic similarity between words. Three new Wiki-based methods, WikiWalk, conceptual graph, and temporal semantic analysis are described in details. The future directions of this field continue to combine Wikipedia and other background information as complementary semantic resources. In addition, characterizing relatedness between words by performing a complex network analysis is also a future challenge.

This paper applies the content analysis method combined with visualization tools UCINET to extract the marketing strategies and corporate strategic intelligence from banner advertising of three China telecom operators’ online business hall. The results show that the content analysis method can effectively extract the marketing strategy and corporate strategic intelligence of telecom operators. Moreover, the strategies extracted from advertising are in good agreement with the company’s real strategies. Therefore, the content analysis method is an effective method for extracting business online advertising intelligence.

Currently, the research of patent technology effect matrix and technology application map is primarily depend on the judgement of experts in classification of technology, effect and application terms, which requires many labors and the result seems to be not comprehensive. Therefore,the paper proposes a relatively objective solution based on text mining. This solution mainly focuses on the extraction of patent information about technology, effect and application terms, and then the experts make decisions based on the results of extraction to define the relatively comprehensive and accurate terms. With these terms, the maps by calculating the numbers of terms and the times of occurrences among them can be built. Experimental results show that the finally statistics are more comprehensive relative to experts’ ways.

This paper proposes a new key indicators definition method which is key ratio. The key ratio gives a field vision of statistics, and the data standardization method and maximum have impacts to key ratio. The key ratio may affect indictor weight, and the method of key ratio is suitable for linear and nonlinear evaluation method.

This paper presents a method for identifying candidate reference control gene based on text mining from PubMed database. It integrates several approaches such as pattern matching, subject recognition and information extraction to find candidate gene and its experiment environment for biology domain specialists. Experiment results show that the method not only has good performance on mining of candidate reference control gene and its environments, but also saves much time and reduces cost.

Shanghai Library chooses the QR code as the standard to generate the two-dimensional code including all the common information of Shanghai Central Library members,then prints them on the Shanghai Central Library One-card-through map and uses Shanghai Library’s smart-phone app or other apps to identify all the information, especially the GEO information inside the QR code.Combined with navigation apps, the users can get to the library more efficiently.

This paper introduces the current application status of wireless LAN in library, analyzes the necessity of the use of open source software to develop WLAN authentication system, then describes the wireless network authentication system design and implementation process, and summarizes the use of WLAN authentication system.

Base on AIMLBot intelligent robot software, the massive intelligent segmentation and Lucene.net are used to implement word segmentation and index retrieval for Chinese knowledge base,while the questions similarity between the author submitting and from the knowledge base are computed by HowNet, which lead to a highest similarity question for reasoning. At last,robot in real-time virtual reference service of the library is implemented.

Based on application analysis of database of present integrated library management system, this paper designs a framework for managing database of integrated library system with Oracle RAC cluster for Shanghai Jiaotong University. For the purperse of optimization of database system and hardware, the procedure of implement on database of integrated library management system is also described.