基于领域本体和概念向量的中文文本相似性测度研究   总被引：2，自引：0，他引：2  

颜端武  成晓  甘利人 《中国图书馆学报》2007,33(6)

文本相似性测度被广泛用于计算用户提问与文档资源相关程度以及基于内容相似资源推荐。OCVSM是一种基于领域本体和概念向量相似性测度的方法。该方法将军用飞机领域知识本体OntoAvion的概念集作为词汇抽取特征项,根据本体中概念间的关系确定特征项的相似度,最后利用余弦算法计算文本向量相似度。实验证明,该方法与基于语言学词典的相似性测度方法相比,更接近用户对文本相似性的判断。表10。图5。参考文献10。  相似文献   

网络文本评论中产品特征抽取综述     

李光敏  陈炽  邢江  彭绪富  杨怿 《现代情报》2016,36(8):168-173

随着Web2.0技术和电子商务的飞速发展,越来越多的用户通过互联网分享产品的使用体验和表达喜恶的观点,如何有效地从评论文本中抽取产品特征是产品意见挖掘方向所亟需解决的问题。本文从产品特征频繁项、特征-意见共现关系、模型训练和显隐式特征匹配四个方面介绍产品特征抽取工作在国内外的研究进展并指出其各自的优势和不足,最后列出产品特征抽取在今后的研究方向。  相似文献   

Inference and evaluation of the multinomial mixture model for text clustering     

Loïs Rigouste  Olivier Cappé  François Yvon 《Information processing & management》2007

In this article, we investigate the use of a probabilistic model for unsupervised clustering in text collections. Unsupervised clustering has become a basic module for many intelligent text processing applications, such as information retrieval, text classification or information extraction.  相似文献   

基于模式聚合和决策树的文本分类规则抽取     

王煜  王正欧 《情报科学》2006,24(1):96-99,123

本文首先提出一种改进的X^2统计量，以此衡量词条对文本分类的贡献。然后根据模式聚合理论，将对各文本类分类贡献比例相近似的词条聚合为一个特征，建立出文本集的特征向量空间模型。此方法有效地降低了文本特征向量空间的维数。最后使用决策树进行分类，从而既保证了分类精度又获得了决策树易于抽取可理解的分类规则的优势。  相似文献   

Patent document categorization based on semantic structural information     

Jae-Ho Kim  Key-Sun Choi 《Information processing & management》2007

The number of patent documents is currently rising rapidly worldwide, creating the need for an automatic categorization system to replace time-consuming and labor-intensive manual categorization. Because accurate patent classification is crucial to search for relevant existing patents in a certain field, patent categorization is a very important and useful field. As patent documents are structural documents with their own characteristics distinguished from general documents, these unique traits should be considered in the patent categorization process. In this paper, we categorize Japanese patent documents automatically, focusing on their characteristics: patents are structured by claims, purposes, effects, embodiments of the invention, and so on. We propose a patent document categorization method that uses the k-NN (k-Nearest Neighbour) approach. In order to retrieve similar documents from a training document set, some specific components to denote the so-called semantic elements, such as claim, purpose, and application field, are compared instead of the whole texts. Because those specific components are identified by various user-defined tags, first all of the components are clustered into several semantic elements. Such semantically clustered structural components are the basic features of patent categorization. We can achieve a 74% improvement of categorization performance over a baseline system that does not use the structural information of the patent.  相似文献   

膜萃取和流动注射在线样品前处理技术   总被引：1，自引：0，他引：1  

刘景富  江桂斌 《中国科学院研究生院学报》2003,20(4):504-509

对用膜萃取和流动注射技术进行样品的在线预处理方法进行了较为系统的研究。第一部分提出了一种新的样品前处理技术———连续流动液膜萃取(CFLME),并用于极性有机污染物的痕量富集。阐述了CFLME的基本原理,并以磺酰脲类除草剂和内分泌干扰物双酚 A为模型化合物,研究了CFLME的影响因素。在优化的条件下,甲磺隆和双酚 A分别经过 1 2 0min和 40min的萃取后,可达到 1 0 0 0倍和2 0 0倍的富集倍数,富集效率是SLM的 3.5～ 2 0 0倍。在此基础上,建立了CFLME 高效液相色谱(HPLC)在线联用测定磺酰脲类除草剂的方法,样品富集 1 0min就能达到 0.0 5～ 0.1 μg/L的检测限;建立的CFLME C1 8预柱 HPLC/紫外检测器在线联用系统,可测定地表水中ng/L级的磺酰脲类除草剂,其检测限比用C1 8SPE柱富集、HPLC/紫外检测器测定时低 2 0 0倍。研究表明,CFLME的主要优点是高选择性、高富集效率、低成本、液膜长期稳定,并易与各种分析仪器联用。第二部分研究了几种流动注射(FI)在线样品前处理技术,为FI应用于日常分析提供了新的思路和途径。建立了FI微孔膜液萃取(MM LLE)系统,用于洗涤剂中阴离子表面活性剂的日常测定;发展了FI高温反应系统,并成功地用于洗涤剂中的总无机磷酸盐以及烟草中总还原糖的自动分析;提出用试剂注入 FI技术  相似文献   

Warning readers to avoid irrelevant information: When being vague might be valuable     

Annie Peshkam  Michael C. Mensink 《Contemporary educational psychology》2011,36(3):219-231

Students are often provided with instructions that are intended to influence their attention to particular sections or elements of their reading materials. To date, the bulk of the work on such prereading instructions has focused on drawing reader attention to relevant text information. In the current project, we examined whether instructions might also be useful in helping readers ignore irrelevant (albeit inherently interesting) information in text. In two experiments, prereading instructions asked readers to (a) focus on specific relevant text segments, (b) ignore specific irrelevant text segments, (c) maintain an awareness that the text contained irrelevant segments without specifically identifying them, or (d) read without warnings. Participants generally exhibited longer reading times and enhanced recall for irrelevant segments compared to base content, except in cases for which general instructions warned about but did not specifically identify those irrelevant elements. The implications of these findings for research on seductive details and text processing, as well practical applications for the design of reading instruction, are discussed.  相似文献   

局部分层非负张量分解算法及在机械故障诊断中的应用     

王飞  许飞云  王海军 《东南大学学报》2011,(4):394-399

针对传统非负张量分解收敛速度慢,分解精度低的难题,结合three semi-NMF模型,将局部目标函数理论应用于非负张量分解中,提出了基于局部分层的非负张量分解算法,并通过人脸特征提取实验验证了算法的有效性.通过对由空压机不同故障振动信号的双谱构成的张量按该算法分解,得到反映故障特征的基图像及与基图像对应的权值矩阵,建...  相似文献   

文本与数据挖掘的著作权困境及应对   总被引：1，自引：0，他引：1  

王文敏 《图书馆理论与实践》2020,(3):28-34

近年来,文本和数据挖掘广泛应用于多个领域,在推动社会经济快速发展的同时也极大地冲击着现有的著作权制度。文章分析了文本和数据挖掘面临的法律困境,并基于国际上两大法系对文本和数据挖掘的立法和司法实践,为我国解决文本和数据挖掘引发的法律困境提出建议。  相似文献   

Assessing books’ depth and breadth via multi-level mining on tables of contents     

《Journal of Informetrics》2020,14(2):101032

Compared with journal articles, books can provide broader, deeper and more comprehensive information, and often have higher expertise and academic depth. However, most researches on book assessment focus on measuring academic value of books (e.g. citations analysis) or identifying attitudes of readers (e.g. book review mining), depth and breadth reflected by book contents is neglected. Therefore, in this paper, we measure books’ depth and breadth by mining books’ tables of contents, so as to enrich resources and methods for book assessment research, help users understand book contents quickly and improve efficiency of book selection. Specifically, we measured books’ depth and breadth based on books’ tables of contents via two levels: topic level and feature level. Firstly, we obtained topic-level metrics by identifying topics expressed in tables of contents and calculating topic distributions. Then, we got feature-level results via feature extraction and feature distribution calculation. Finally, we compared depth and breadth metrics and other book assessment metrics. Experimental results reveal that, books’ depth and breadth at two levels are different, and substantial differences between disciplines and book types are obvious. In addition, books’ depth and breadth can provide alternative and supplementary information for assessing multi-dimensional values of books.  相似文献