Extending WHIRL with background knowledge for improved text classification     

Sarah Zelikovitz  William W. Cohen  Haym Hirsh 《Information Retrieval》2007,10(1):35-67

Intelligent use of the many diverse forms of data available on the Internet requires new tools for managing and manipulating heterogeneous forms of information. This paper uses WHIRL, an extension of relational databases that can manipulate textual data using statistical similarity measures developed by the information retrieval community. We show that although WHIRL is designed for more general similarity-based reasoning tasks, it is competitive with mature systems designed explicitly for inductive classification. In particular, WHIRL is well suited for combining different sources of knowledge in the classification process. We show on a diverse set of tasks that the use of appropriate sets of unlabeled background knowledge often decreases error rates, particularly if the number of examples or the size of the strings in the training set is small. This is especially useful when labeling text is a labor-intensive job and when there is a large amount of information available about a particular problem on the World Wide Web.

Haym HirshEmail:

  相似文献   

两个加工时间恶化的半线性平行机排序问题的启发式MBLS 算法     

程明宝  孙世杰 《上海大学学报(英文版)》2007,11(5):451-456

The combination of online or semi-online with deterioration jobs has never been researched in scheduling problems. In this paper, two semi-online parallel machine scheduling problems with linear deterioration processing time are considered. In the first problem, it is assumed that the deterioration rates of jobs are known in an interval, that is, bj ∈[0, α], where 0 〈α≤ 1 and bj denotes the linear deterioration rate. In the second problem, it is assumed that the largest deterioration rate of jobs is known in advance, that is, b = max1≤j≤n {bj }. For each of the two problems, a heuristic MBLS algorithm is worked out and its worst-case ratio is analyzed. At the same time, the worst-case ratio of the list （LS） algorithm is investigated and it is proved that all the ratios are tight.  相似文献   

Effect of lead on ERK activity and the protective function of bFGF in rat primary culture astroglia     

Zhang Y  Ye LP  Wang B  Cao SC  Sun LG 《Journal of Zhejiang University. Science. B》2007,8(6):422-427

  相似文献   

五羊-本田市场营销策略的博弈分析及启示     

尚万春 《安阳工学院学报》2007,(2):69-71

针对五羊-本田市场营销策略的现状:高品质、高成本、高价格、低利润、市场萎缩,以价格作为切入点,用博弈论中不完全信息静态博弈、不完全信息动态博弈的方法浅析其营销策略的有效性,并得出五羊-本田市场营销策略不妥的结论及对企业在制定营销策略中的启示。  相似文献   

Phonological processing and emergent literacy in younger and older preschool children     

Anthony JL  Williams JM  McDonald R  Francis DJ 《Annals of dyslexia》2007,57(2):113-137

Phonological awareness, phonological memory, and phonological access to lexical storage play important roles in acquiring literacy. We examined the convergent, discriminant, and predictive validity of these phonological processing abilities (PPA) in 389 3-, 4-, and 5-year-old children. Confirmatory factor analysis supported the validity of each PPA as separate from general cognitive ability and separate from each other. Multigroup structural equation modeling (SEM) with mean structure demonstrated that older preschoolers have better developed latent PPA than younger preschoolers but that the structure of PPA is equivalent. RAN was found uniquely associated with letter knowledge and text discrimination in younger preschoolers, and PA was found uniquely associated with word reading skills in older preschoolers. Finally, general cognitive ability was only indirectly associated with emergent literacy via PPA. These results highlight the importance of PPA in the early literacy development of English-speaking preschool children.  相似文献   

Evaluation and publication delay in Ibero-American scientific journals     

Jon Zabala  Borja González-Albo  Ana García-García  Aurora Garrido-Domínguez  José Ignacio Vidal-Liy  Luis R. Álvarez-Díez  Soledad Hernando-Tundidor  Yara Mostazo-Fernández  Teresa Abejón 《Learned Publishing》2023,36(2):205-216

This article analyses the review, acceptance and publication dates of a sample of 21,890 articles from 326 Ibero-American scientific journals from all subject areas and countries included in the Latindex Catalogue 2.0 and published between 2018 and 2020 (freely available as an open access dataset). The aim is to discover evaluation and publication times. The evaluation process takes a median of 110 days, the publication process, a median of 82 days, and the whole process, a median of 224 days. Statistical differences are found according to periodicity, subject areas, countries, existence of a printed version and article type (Call for Papers or General articles). From the data we find that the delay in publication is longer than publishers themselves report to the DOAJ. STEM areas present the most similarity in publication patterns, having a higher number of evaluation days (Ed) than publication days (Pd); Arts and Humanities present the opposite pattern, with a higher Pd than Ed. In the case of Social Sciences, the times are similar. General articles and Call for Papers articles differ in terms of Ed, but not Pd, indicating that Call for Papers revisions are faster.  相似文献   

使用善锋软件审核加工参考文献的几点思考     

周凤航  金铁成 《编辑学报》2023,(4):426-428

善锋软件是编辑审核参考文献的有效工具。本文总结了善锋软件的诸多实用功能:定制参考文献著录内容和著录格式,标记出新、旧参考文献的所有差异并协助编辑纠正各类文字性和格式性差错,提供参考文献原文链接,对参考文献的遗漏、颠倒、重复等问题进行校验,对文内、文后参考文献责任者姓名与序号(出版年份)一致性进行检验等。来源数据的先天质量问题可能导致善锋软件输出结果出现对应的细节性偏差。使用善锋软件时,只要严格执行使用流程,对软件标记出的疑似差错进行原文核查,就可以消除参考文献中的各类差错。  相似文献   

固体火箭发动机试验数据处理软件设计     

冯喜平  董韬  李进贤  曹琪  李洁 《实验技术与管理》2011,28(6):33-36

针对火箭发动机地面试验中的数据处理需求,按照火箭发动机试验数据处理规范(通用产品)和火箭发动机试验数据处理方法(特定产品),使用LabVIEW7.1开发了发动机试验数据处理软件,实现了发动机试验数据的快速处理和试验报告的自动生成,满足了火箭发动机试验后数据处理要求。软件已成功应用于多种发动机产品试验数据处理中。成果获得2 0 0 9年度中国人民解放军总装备部科技进步三等奖。  相似文献   

集成运放电路的信号流图分析     

刘树平  吴勇灵 《黔南民族师范学院学报》2011,(6):29-32,79

一个线性电网络存在着由其线性方程组确定的信号流图。而线性方程组的代数变换与图的变换存在着对应关系,通过图的变换可直接对一个线性电网络求解。根据这一原理,本文论述了信号流图在集成运放电路分析中的应用,并通过两个实例加以说明。  相似文献   

Length-based vehicle classification in multi-lane traffic flow     

于洋  于明  阎刚  翟艳东 《天津大学学报(英文版)》2011,17(5):362-368

For the realtime classification of moving vehicles in the multi-lane traffic video sequences, a length-based method is proposed. To extract the moving regions of interest, the difference image between the updated background and current frame is obtained by using background subtraction, and then an edge-based shadow removal algorithm is implemented. Moreover, a thresholding segmentation method for the region detection of moving vehicle based on location search is developed. At the estimation stage, a registration line is set up in the detection area, then the vehicle length is estimated with the horizontal projection technique as soon as the vehicle leaves the registration line. Lastly, the vehicle is classified according to its length and the classification threshold. The proposed method is different from traditional methods that require complex camera calibrations. It calculates the pixel-based vehicle length by using uncalibrated traffic video sequences at lower computational cost. Furthermore, only one registration line is set up, which has high flexibility. Experimental results of three traffic video sequences show that the classification accuracies for the large and small vehicles are 97.1% and 96.7% respectively, which demonstrates the effectiveness of the proposed method.  相似文献