Current Issue
    , Volume 29 Issue 1 Previous Issue    Next Issue
    For Selected: View Abstracts Toggle Thumbnails
    Research on Representative Semantic Models for Linking and Organizing Digital Objects
    Ma Yumeng, Zhu Zhongming
    2013, 29 (1): 1-7.  DOI: 10.11925/infotech.1003-3513.2013.01.01
    Abstract   HTML   PDF (759KB) ( 381 )
    In the linked open environment driven by semantic and data, digital repositories, digital libraries and other domains need semantic digital objects and their linked relations to support semantic computing, making resources semantic linked and organized. This paper analyses Fedora, EDM, CERIF and Research Object which are representative semantic models for linking and organizing digital objects, and makes a comparative analysis and recommendation of model selection.
    References | Related Articles | Metrics
    Identifying Query Intent by Exploiting Query Refinement
    Zhang Xiaojuan, Lu Wei
    2013, 29 (1): 8-14.  DOI: 10.11925/infotech.1003-3513.2013.01.02
    Abstract   HTML   PDF (714KB) ( 343 )
    Based on the AOL log dataset, this paper tries to exploit query reformation to identify the concrete query intent of users without given query intent category system. This paper mainly discusses how to identify the query reformation which can express the user intent of original query and how to cluster the query intent. The final results evaluated manually show that this experiment achieves a good effect.
    References | Related Articles | Metrics
    The Study on Out-of-vocabulary Identification of Chinese Biomedical Field Based on Hybrid Method
    Sun Haixia, Li Junlian, Wu Yingjie, Wu Suhui
    2013, 29 (1): 15-21.  DOI: 10.11925/infotech.1003-3513.2013.01.03
    Abstract   HTML   PDF (1057KB) ( 324 )
    First, the status of research on out-of-vocabulary automatic identification is introduced briefly. Then,combining the word length distribution and morphological characteristics of Chinese biomedical field, this paper presents an hybrid method of out-of-vocabulary identification of Chinese biomedical field, which is based on N-gram, integrating the methods of the field dictionary-based, filtered corpus-based, and rules-based. Finally, on a sample set of pharmaceutical journals data of Chinese BioMedical Literature Database, the authors make an experiment of the proposed hybrid method, and the experimental results achieve a good performance.
    References | Related Articles | Metrics
    Combining Logical Inference with Content-based Computing for Intelligent Retrieval in Academical Networks
    Nie Hui
    2013, 29 (1): 22-29.  DOI: 10.11925/infotech.1003-3513.2013.01.04
    Abstract   HTML   PDF (1338KB) ( 305 )
    The expression ability of Ontology description language OWL-DL is restricted in description logic. The actual utilization regarding Ontology is impacted due to the implicated relations among Ontology individuals not being able to be detected. With regard to the issue, the SWRL-based inference mechanism for knowledge base is introduced, by which semantic relations implied in the knowledge base can be identified. Consequently, implicit knowledge is embodied explicitly and more extensive inference results can be available. The mechanism is employed to tackle the problem of implicit knowledge discovery of academic resources on the Web. Furthermore, the topic-specific relations for the academic resources are built on the basis of the content-based similarity measure. All regarding approaches are tested in the prototype indicating reasonability, feasibility and effectiveness of the scheme.
    References | Related Articles | Metrics
    Research on Collaborative Filtering of Heuristic Transitive Similarity Between Items
    Li Linna, Li Jianchun, Zhang Zhiping
    2013, 29 (1): 30-35.  DOI: 10.11925/infotech.1003-3513.2013.01.05
    Abstract   HTML   PDF (874KB) ( 417 )
    Aiming at the problem of only finding similar relationship between items rated by common users and enlightened by the transitivity between peoples among social network, this paper figures that the similarity between items also have transitivity. A collaborative filtering algorithm based on heuristic similarity propagation between items is proposed. The experiments indicate that the proposed method can provide better recommendation accuracy by comparing with classic collaborative filtering algorithms.
    References | Related Articles | Metrics
    Research on Three-dimensional Personalized Recommendation Approach for C2C E-commerce Platform
    Ai Danxiang, Zuo Hui, Yang Jun
    2013, 29 (1): 36-42.  DOI: 10.11925/infotech.1003-3513.2013.01.06
    Abstract   HTML   PDF (886KB) ( 312 )
    This paper defines a three-dimensional recommendation space and recommendation task in C2C e-commerce platforms, which are different from B2C ones, and proposes a three-dimensional personalized recommendation approach, which extends the traditional two-dimensional collaborative filtering method and content-based recommendation method. The proposed approach firstly calculates seller similarities using seller features, and fills the three-dimensional rating set based on sales relations and seller similarities to solve the data sparsity problem. Then it calculates buyer similarities using historical ratings to decide neighbors and predict unknown ratings. A true data experiment proves that the proposed approach is effective to solve the personalized recommendation problem in C2C platforms and has good performance when recommending seller and product combinations.
    References | Related Articles | Metrics
    Mechanical Design Image Retrieval with Combined Geometrical Features
    Fang Naiwei, Lv Xueqiang, Zhang Dan
    2013, 29 (1): 43-49.  DOI: 10.11925/infotech.1003-3513.2013.01.07
    Abstract   HTML   PDF (852KB) ( 374 )
    Content-based mechanical design image retrieval is of great importance for the mechanical design industry. According to the general characteristics of mechanical design images, this paper proposes a new retrieval method based on combined geometrical features. Firstly, seven features such as solidity, rectangular degree and so on, are extracted from the shape region of a mechanical design image, all the features easily obtained through computing the perimeter, area,etc, and need no normalization. Secondly, the selected features are combined to determine the shape feature descriptor. The proposed descriptor is then applied in mechanical design image retrieval. The experiments show that the proposed method performs better than Fourier Descriptors and Hu invariant moments in the retrieval of mechanical design images.
    References | Related Articles | Metrics
    Study on the Keyword Extraction from Roadmap Based on the Lexical Chains
    Ye Chunlei, Leng Fuhai
    2013, 29 (1): 50-56.  DOI: 10.11925/infotech.1003-3513.2013.01.08
    Abstract   HTML   PDF (576KB) ( 384 )
    The paper proposes a method to extract the keyword based on the lexical chains. The method can describe the technical field topics in the technology roadmap by constructing lexical chains, and regard the lexical chains as semantic relations of keyword in the technical field. The experiment shows that this method can extract the keyword to reveal the content of technical field in technology roadmap more comprehensively, and can significantly improve the precision and recall rate than TF-IDF.
    References | Related Articles | Metrics
    On the Scientific Research Teams Identification Method Taking Co-authorship of Collaboration as the Source Data
    Shen Gengyu, Huang Shuiqing, Wang Dongbo
    2013, 29 (1): 57-62.  DOI: 10.11925/infotech.1003-3513.2013.01.09
    Abstract   HTML   PDF (675KB) ( 386 )
    In the research on personal and institutional evaluation, it is difficult to guarantee the reliability and accuracy of identifying the scientific research team.This paper applies the vector space model into the identification of scientific research teams within the co-authorship network. Under the premise of considering the authorship order in the paper, and by constructing the vector space of papers and authors, the collaboration relationship is measured by calculating the degree of similarity between author vectors. Then,this paper analyses the collaboration network with the analytical approach of cohesion sub-group in social network analysis. At last,by choosing all its faculty of a department in a university as research object,the present research accurately identifies all the scientific research teams that exist in the institution and the rationality of this method is verified.
    References | Related Articles | Metrics
    Research on Review Spam Recognition
    Li Xiao, Ding Shengchun
    2013, 29 (1): 63-68.  DOI: 10.11925/infotech.1003-3513.2013.01.10
    Abstract   HTML   PDF (510KB) ( 529 )
    This paper analyses review spam from the perspective of the usefulness of information, selects digital camera reviews as the research object and builds the data set, then from the three aspects of review, reviewer and product chooses 11 features, uses 4 different kernel functions in SVM model to identify review spam of products, optimizes the parameters C and γ of RBF that has a better identification, which improves accuracy rate of the identification effect of review spam to 78.16% and recall rate to 72.18%. By comparing the selected 4 different combinations of features, the authors find the combination of review, reviewer and product is the best. Finally, it proves that SVM is significantly better than other algorithms compared to the Logistic Regression.
    References | Related Articles | Metrics
    Research on Information Quality and Credibility Evaluation in Online Community——Based on User Perspective
    Shen Wang, Guo Jia, Li He
    2013, 29 (1): 69-74.  DOI: 10.11925/infotech.1003-3513.2013.01.11
    Abstract   HTML   PDF (548KB) ( 334 )
    This paper adopts Grounded theory coding methods to research on user-generated information quality and information credibility evaluation in the context of online communities,and analyzes a total of 13 020 messages of two discussion topics.16.8% of the messages have clear evaluation of information quality and credibility. The results show that:(I)the users usually use one or a few evaluation indicators to evaluate information quality and credibility in online community; (II)the user frequently use correctness, usefulness, specificity to evaluate information quality and reputation, professionalism, honesty (negative standard) to evaluate information credibility;(III) the users tend to use negative indicators to evaluate the quality and reliability of online community information.
    References | Related Articles | Metrics
    Research on Consumer Satisfaction Index Evaluation Model of Online Resources for National Elaborate Curriculum
    Hu Dehua, Ren Lei, Che Dan
    2013, 29 (1): 75-82.  DOI: 10.11925/infotech.1003-3513.2013.01.12
    Abstract   HTML   PDF (843KB) ( 437 )
    Consumer Satisfaction Index (CSI) evaluation model of online resources is built for National Elaborate Curriculum(NEC), which has seven latent variables such as user experience, user need, user expect, perceived quality, perceived value, user satisfaction and user loyalty. And then the authors conduct an empirical research on the model by Smart PLS and confirm the scientific of the model. In the end, the application of the model is evaluated and analyzed.
    References | Related Articles | Metrics
    Design and Implementation of Distributed Collaborative Filtering Algorithm on Hadoop
    Xiao Qiang, Zhu Qinghua, Zheng Hua, Wu Kewen
    2013, 29 (1): 83-89.  DOI: 10.11925/infotech.1003-3513.2013.01.13
    Abstract   HTML   PDF (598KB) ( 763 )
    Based on Hadoop, this paper demonstrates that traditional collaborative filtering algorithm cannot adjust to cloud computing platform, then improves traditional collaborative filtering algorithm to adapt to the Hadoop platform from similarity and prediction,and also achieves sequential modular MapReduce collaborative filtering computing tasks.
    References | Related Articles | Metrics
    Integration of Huiwen OPAC and E-reading——Take Library of Hunan Institute of Science and Technology for Example
    Zhao Lin, Hu Jianhong
    2013, 29 (1): 90-94.  DOI: 10.11925/infotech.1003-3513.2013.01.14
    Abstract   HTML   PDF (847KB) ( 395 )
    According to the 3rd period of CALIS and using the technologies of PHP,XML,JSON,jQuery, the paper integrates CALIS's e-reading into the OPAC of Library of Hunan Institute of Science and Technology based on Mashup, which realize the detail information page of book catalogue shows the full text and try-read of e-books, catalog, cover and so on without any page refresh, as to enrich the OPAC's content, promote user friendly experience and service ability of the library.
    References | Related Articles | Metrics
  Copyright © 2016 Data Analysis and Knowledge Discovery   Tel/Fax:(010)82626611-6626,82624938   E-mail:jishu@mail.las.ac.cn