RDF linkage should be solved with the development of linked data from publishing toward gradually integration application. This paper analyses approaches for RDF linkage algorithm such as support vector machine, backlink based on HTTP referer, schema mapping and R2R property transformation. In addition, linkage frameworks such as Silk, LIMES and R2R are explored from the angle of system framework, language and key algorithm. Finally, future development of RDF linkage are discussed in the perspective of linkage algorithm, framework and application.

In this paper, ACRL (2000), ANZIIL (2004), SCONUL (2011) and Information Literacy Competency Standards for Higher Education in Beijing(Beijing(2006)) are compared and the conclusion that ACRL (2000) is the most comprehensive of them is get. With the IT skills considered in SCONUL (2011) and Beijing(2006), the model is built based on ACRL (2000) and computer self-efficacy. 929 copies of questionnaire are analyzed by Structural Equation Model(SEM),and results reveal that (Ⅰ) ACRL (2000) is appropriate to evaluate information literacy; (Ⅱ) Computer self-efficacy does not belong to information literacy competency; (Ⅲ) The path coefficients indicate that information evaluation, information need, and information ethics are the main element of information literacy, meanwhile information access and information use are relatively less significant.

Due to the current Ontology construction problems using Folksonomy, the limitations of the current approaches are summarized, and the advantages of FCA’s characteristics on solving these defects are pointed out,because it has the functions of data analysis and conceptual modeling rolled into one.By the way of building the Ontology construction model, a new Ontology construction modeling method based on FCA and Folksonomy is proposed, and it provides some new ideas and perspectives for achieving a low-cost, efficient, timely, flexible and people-oriented Ontology building process in Web community environment.

In analyzing unstructured information of patents, there is a problem in identifying and defining the technology innovations and the effect of patent currently.This paper puts forward a method to differentiate technical and effect words in patent,based on the features of patents’ structure-grammar-clue word.The method can synthetically consider three feature factors: the structure, the grammar and clues word, then improve the recognition result of the technical and effect words in patents.

This paper implements String Frequency Statistics Algorithm proposed by Nagao to build Proper Noun Recognition (PNR) system for Chinese and Bengali languages. First, n-grams are extracted from untagged input corpus,then they are filtered to get rid of redundant sub-strings, using SSR algorithm. Finally, this multilingual PNR system assigns each n-gram a probability of being a proper noun based on the information of their neighboring words and outputs results according to their probability score. The test results show that this system can effectively recognize name of people, places, organizations or institutions from the input text.

Based on the whole Chinese booklist of a certain university library as well as the analysis of its book indexing information, the paper summarizes the features and extracting laws of Chinese keywords, and establishes a Chinese keywords extraction model based on characters sequence annotation, which proposes the basic idea and implementation scheme for extracting keywords. It verifies the feasibility, rationality and practicality of the model by large-scale experiments, and basically solves the problems of Chinese keywords extraction without executing words segmentation, which shows that characters sequence annotation is better than words sequence annotation.

This paper proposes a method,which clusters the initial terms collection by ant colony algorithm and clusters the results hierarchy by K-means algorithm, then gets the labels of classes using the comprehensive similarity calculation, finishes the term hierarchy relation’s structure at last. Parts of experimental results are appraised and analyzed by domain experts.

In order to get dynamics of Web information timely, an online Web topic clustering approach based on Single-Pass algorithm is researched. The clustering process of this approach is analyzed firstly,and the key problems including extracting and weight calculating of features as well as representation and modification of topic cluster are deliberated. Experiment is designed to compare the effects of different weight factor of features in title, weight calculating and normalizing methods of features and the vector dimension of topic cluster on cluster quality and time efficiency.

In order to reduce the poor consistency and the errors in manual patent classification, this article introduces text classification technology into patent auto-classification system. It uses the radial basis function neural network algorithm to realize the automatic classification of patent text, and analyses the test samples.The experiment results show that this new system has a better classification results,and the average F1 value is higher than 70%.

Based on the theory of text mining, this paper puts forward a contrast method of product reviews in different shopping websites, and makes analysis on whether the product reviews from different shopping websites are consistent. Firstly,this paper analyses the reviews of product feature one by one. Then, it makes contrast analysis from one product feature to total product features. The study discovers that the reviews of the same product from different shopping websites are not completely consistent, and this inconsistency mainly reflects in product features, which means product reviews will be different due to different shopping websites.

Grounded on the R open source statistic environment, this paper builds a data mining model on the complex HPLC multi-dimensional fingerprint data with visualization analysis. The result fully reflects that PCA can be used as a model to reveal the principles of multi-dimensional data, and verifies the validity and practicality of principal component and cluster analysis and neural network to reveal the character of multi-dimensional data. Finally, based on the results of PCA and cluster analysis, this paper builds a training network model through techniques of machine learning and other related statistical algorithm to predict the habitat of unknown TCM sample, which supplies sufficient evidences to the TCM quality control.

This paper investigates user’s searching behaviour of library resource discovery service by log mining. First of all, an intermediate platform is built to generate and record user’s searching logs. Then the log information is identified, extracted and objectified by Java application, and the data is permanently stored in a MySQL Database for further statistics and analysis. The analysis includes searching language, searching length, the using frequency of advanced-search, facets, how users modify their search queries, etc, and the conclusions are helpful to the implementation, application and optimization of library resource discovery service.

In this article, the library of Guangzhou University is success in building a VPN system by using RADIUS remote authentication, utilizing unified authentication server, and integrating VPN with ILAS III through the user data synchronization system. This VPN system not only can ensure the legitimacy, effectiveness and safety for users, but also can provide convenience for users and system manager.

This article concerns on the security problems faced by computer room of digital library. Because of the tight budget and the lack of working personnel in university library, a kind of low cost scheme about security problems of computer room is studied based on the situation of Changzhou Institute of Technology Library. Then a B/S model computer room remote monitoring system of digital library is designed and implemented based on the PIC microcontroller and wireless communication. The experiment results show that the system is stable and reliable,and the scheme and system have some referential value to current safety construction of computer room in digital library.