Hierarchical clustering of a Finnish newspaper article collection with graded relevance assessments     

Tuomo Korenius  Jorma Laurikkala  Martti Juhola  Kalervo Järvelin 《Information Retrieval》2006,9(1):33-53

Search facilitated with agglomerative hierarchical clustering methods was studied in a collection of Finnish newspaper articles (N = 53,893). To allow quick experiments, clustering was applied to a sample (N = 5,000) that was reduced with principal components analysis. The dendrograms were heuristically cut to find an optimal partition, whose clusters were compared with each of the 30 queries to retrieve the best-matching cluster. The four-level relevance assessment was collapsed into a binary one by (A) considering all the relevant and (B) only the highly relevant documents relevant, respectively. Single linkage (SL) was the worst method. It created many tiny clusters, and, consequently, searches enabled with it had high precision and low recall. The complete linkage (CL), average linkage (AL), and Ward's methods (WM) returned reasonably-sized clusters typically of 18–32 documents. Their recall (A: 27–52%, B: 50–82%) and precision (A: 83–90%, B: 18–21%) was higher than and comparable to those of the SL clusters, respectively. The AL and WM clustering had 1–8% better effectiveness than nearest neighbor searching (NN), and SL and CL were 1–9% less efficient that NN. However, the differences were statistically insignificant. When evaluated with the liberal assessment A, the results suggest that the AL and WM clustering offer better retrieval ability than NN. Assessment B renders the AL and WM clustering better than NN, when recall is considered more important than precision. The results imply that collections in the highly inflectional and agglutinative languages, such as Finnish, may be clustered as the collections in English, provided that documents are appropriately preprocessed.  相似文献   

从信息构建看未来的知识管理   总被引：4，自引：0，他引：4  

化柏林 《情报学报》2004,23(4):463-468

本文从信息构建出发 ,进而引申到知识管理与知识推理 ,最终提出了知识基础工程。在设计并实现自然语言的语法开发平台时 ,产生了构建知识库的设想。利用语法开发平台 ,就可以对自然语言的语法进行开发 ,加上一个好的算法 ,就可以对大规模文本进行自动分析。对分析过的句子进行内容提取 ,并用面向对象方法和逻辑形式进行格式化 ,得到以面向对象为特征的常识知识库和以逻辑命题为特征的专家知识库 ,这应该是知识发现、知识管理的最高层次 ,也是知识工程的核心  相似文献   

网上外文科技期刊最新论文题录保障系统的研制     

骆达荣 《图书情报工作》2004,48(3):93-94

介绍网上外文科技期刊最新论文题录保障系统的研制目标、总体结构及主要功能,着重阐述系统自动下载和自动标 引的实现方法。  相似文献   

试论多媒体信息检索的发展     

丁大可  李树青 《图书馆论坛》2004,24(6):204-206,272

文章通过对文本信息检索语言发展的分析，得出检索语言发展的一般规律，由此对多媒体信息检索语言的发展做出回顾和总结，同时对于多媒体信息检索语言的前景进行了展望。  相似文献   

关于自然语言检索问题   总被引：13，自引：0，他引：13  

张琪玉 《图书馆论坛》2004,24(6):211-213,145

自然语言检索当前仅在关键词检索的层次上已经实现，重点分析了全文检索所适应和不适应的检索要求及其评价，用分析说明了当前关于自然语言检索的主流观点的不切实际，指出自然语言检索的发展方向是与情报检索语言结合和融合。  相似文献   

特殊合订图书的标引与著录     

邓福泉 《图书馆杂志》2004,23(1):42-43

针对特殊合订图书的特点，文章对特殊合订图书的分类标引、主题标引和CN—MARC著录方法分别进行了探讨。  相似文献   

从五十年馆藏西文文献看国外藏学研究及藏学文献的发展   总被引：1，自引：0，他引：1  

董印红 《四川图书馆学报》2004,(1):66-68

通过对馆藏西文藏学文献的分析,以点带面,概括出国外对藏学研究及国外出版的藏学文献的特色、意义和今后的发展方向,以及国内在收藏和利用国外有关文献时应注意的问题,从而为藏学研究工作者提供更好的文献资源服务。  相似文献   

高校图书馆外文期刊利用中的问题与对策     

张帆 《四川图书馆学报》2004,(6):51-52

外文期刊利用率低是各高校图书馆普遍存在的问题，文章对造成外文期刊利用率低的原因进行分析，并提出了相应对策。  相似文献   

探析自动索引原理     

衡中青  熊歆 《图书馆论坛》2004,24(1):95-96,78

文章试图对自动索引的方法作一些探讨。首先介绍了自动索引的五种理论模型和六种自动标引方法，随后探讨了自动索引中不可或缺的两个对象——全文索引和数据库索引的原理。  相似文献   

Automatic Alphabet Recognition     

Maayan Geffet  Yair Wiseman  Dror Feitelson 《Information Retrieval》2005,8(1):25-40

The last step of the Information Retrieval process is to display the found documents to the user. However, some difficulties might occur at that point. English texts are usually written in the ASCII standard. Unlike the English language, many languages have different character sets, and do not have one standard. This plurality of standards causes problems, especially in a web environment, where one may download a document with an unknown standard. This paper suggests a purely automatic way of finding the standard which was used by the document writer based on the statistical letters distribution in the language. We developed a vector-space-based method that creates frequencies vectors for each letter of the language and then matches a new document's vectors to the pre-computed templates. The algorithm was applied on various types of corpora in Hebrew, Russian and English, and provides an efficient solution to the stated problem in most cases.  相似文献