支持近似图查询的Why-Not问题解释方法

why-not问题是为查询结果中的缺失元组找到合理的解释。解决数据库查询中的why-not问题不仅能够帮助用户更好地理解查询,而且能够提高数据库的质量和可用性。为了提高图数据库的可用性,提出了支持近似图查询的why-not问题解释方法。该解释方法不仅阐明了为什么why-not问题没有出现在查询结果中,而且给出了一些修改初始查询图的建议,使得why-not问题能够出现在修改后的查询图的查询结果中。该算法分两部分完成:第一部分为候选修改操作生成阶段,首先利用边频率信息提出候选操作集生成基本算法,接着利用图分解操作提出候选操作集生成改进算法,得到修改初始查询图的候选操作集;第二部分基于对查询图修改操作数最少的代价模型,分别采用贪心算法和回溯法选取候选操作,贪心算法设计了合理的贪心函数,回溯法构建了回溯剪枝树,并提出三种剪枝策略执行剪枝操作,最终选取的候选操作集即为支持近似图查询的why-not问题的合理解释。实验表明,该方法可以快速有效地为近似图查询中的why-not问题提供合理解释。     

Oracle 10g中闪回查询的应用

简述了oracle10g中闪回查询的类型,通过列举实例说明了闪回版本查询和闪回事物查询的应用。闪回查询是最基本的闪回功能,直接利用回滚段中的旧数据构造某一刻的一致性数据版本。由于该查询只适合单个表数据恢复,所以对事务中相关的多表数据恢复不适合,无法确保相关数据的参照完整性。     

综合文档语义与用户查询语义的XML关键字检索

为了解决XML关键字查询中语义信息丢失的问题,提出了一种语义相关的关键字检索方法。利用文档的半结构化特点提取文档隐含的语义,利用查询语法捕获用户查询意图,然后根据用户意图查询满足条件的元素,并结合文档语义,由最小最近公共祖先改进为语义相关实体子树集来表达查询结果。实验结果表明,该方法能够有效提高关键字检索结果的查准率。     

多种查询功能的设计与实现

在数据库管理系统中,查询扮演着重要的角色。该文打破了查询设计的一贯思路,利用 Ｏｒａｃｌｅ＊Ｆｏｒｍｓ实现了多种查询功能。这些查询功能不仅能够完成多表相关数据之间的联合查询,同时实现了细表对主表的查询,保证了各表之间的一致性。     

一种基于排序学习方法的查询扩展技术

查询扩展作为一门重要的信息检索技术,是以用户查询为基础,通过一定策略在原始查询中加入一些相关的扩展词,从而使得查询能够更加准确地描述用户信息需求。排序学习方法利用机器学习的知识构造排序模型对数据进行排序,是当前机器学习与信息检索交叉领域的研究热点。该文尝试利用伪相关反馈技术,在查询扩展中引入排序学习算法,从文档集合中提取与扩展词相关的特征,训练针对于扩展词的排序模型,并利用排序模型对新查询的扩展词集合进行重新排序,将排序后的扩展词根据排序得分赋予相应的权重,加入到原始查询中进行二次检索,从而提高信息检索的准确率。在TREC数据集合上的实验结果表明,引入排序学习算法有助于提高伪相关反馈的检索性能。     

基于关键字的RDF聚合查询研究

目前RDF数据上关键字查询转换为结构化语句的算法主要支持对于一般图元素的查询，而无法转换为包含聚合操作的结构化语句。关键字存在大量候选解释，且可能同时匹配聚合操作或图元素，这导致查询中聚合意图的理解非常困难。对此，提出将关键字查询自动转换为可能包含聚合操作的SPARQL语句的算法。算法对SPARQL所支持的聚合操作进行分类，获得关键字与聚合类别的匹配字典，进行关键字映射，计算关键字可能指示聚合意图的概率，确定候选查询解释，并利用模式图获得查询意图，设计意图分数计算方法和查询转换算法，得到对应的查询语句。LUBM和DBLP数据集上的实验验证了算法的有效性和准确性。     

基于远程服务器工作状态的Internet查询优化模型

在Internet数据集成环境中,服务器工作状态在查询费用计算中起着非常重要的作用。本文利用样本查询分类技术来确定服务器的工作状态,并利用相应技术建立相关的服务器工作状态查询费用模型,同时利用所研究的调整函数来估算整个系统查询的开销。该方法可以通过确定服务器工作状态来精确地预计系统的查询费用。     

基于文件存储GIS的SQL解析器的设计与实现

针对利用文件存储的GIS系统不能自动实现查询这一问题，结合SQL-99标准和GIS软件MAPGIS定义的语法规则，运用编译原理中的词法分析、语法分析等理论，实现了将用户输入的普通查询表达式解释为一个标准的SQL语句，并输出查询结果。该SQL解析器已经成功地运用在MAPGIS文件管理的属性检索模块中。     

关系数据库不可用空值的查询与处理

现有的基于关系数据模型的商业数据库采用空值对缺失信息进行建模与处理,然而,单一的空值解释无法体现空值本身的丰富语义。事实上,在相关研究中空值通常被解释为‘值未知’,‘值不可用’以及‘值不存在’等。文中主要研究不可用空值的查询与处理。通过仔细地观察和深刻地理解,分别在传统关系数据库查询和模糊数据库查询中讨论不同语义背景和查询条件下不可用空值的处理和分类。此外,还针对涉及不可用空值的传统关系数据库查询提出选择运算和差运算算法,这些算法使文中的研究更具实用性。     

一种基于语义单元的查询扩展方法

查询扩展技术通过向初始查询请求中加入相似或者相关的词,来减少查询请求与相关文献在表达上的不匹配现象,改善检索性能.本文利用语义单元的语义表达能力和语义单元之间的关系,将与初始查询具有密切语义关系的查询词或短语加入到初始查询请求中,更加全面地表示了用户的查询意愿.算法的时间复杂度为O(L),只与搜索请求的长度L有关,与语义单元表示库的规模无关,这对实时性要求较高的搜索引擎来讲是很实用的.     

一种大类别数分类的神经网络方法

神经网络是一种普遍使用的分类方法。当类别数目较大时,神经网络结构复杂、训练时间激增、分类性能下降。针对这些问题,基于Ｎ分类问题的两种类方法和树型分类器结构,对两分类子网络集进行排序,中给出了一种大类别分类的神经网络阵一结构和快速搜索方法并重点分析了网络阵列的分类性能。理论分析表明,使用网络阵列方法可降低平均分类错误率。该方法还使得网络结构简单灵活,易于扩充,网络的训练时间缩短,仿真实验表明,该方     

Dynamic Restructuring of E-Catalog Communities Based on User Interaction Patterns

Since e-catalogs are dynamic, autonomous, and heterogeneous, the integration of a potentially large number of dynamic e-catalogs is a delicate and time-consuming task. In this paper, we describe the design and the implementation of a system through which existing on-line product catalogs can be integrated, and the resulting integrated catalogs can be continuously adapted and personalized within a dynamic environment. The integration framework originates from a previous project on integration of Web data, called WebFINDIT. Using the framework, we propose a methodology for adaptation of integrated catalogs based on the observation of customers' interaction patterns.     

结合距离分类器的神经网络手写体汉字识别

手写体汉字识别技术中如何解决复杂的大类别识别问题,是汉字识别中的一个难点。该文介绍了基于笔划的手写体汉字特征抽取方法,提出了一种基于预分类的神经网络汉字识别方法,该方法用一个传统的距离分类器先对汉字进行预分类,神经网络根据预分类结果进行有选择的训练和识别,能有效解决神经网络大类别模式识别中的训练和分类问题,学习时间很短,识别效果较理想。     

Interactive Critiquing forCatalog Navigation in E-Commerce

E-commerce sites can have large, essentiallyunbounded, catalogs. With large catalogs comesincreasing difficulty for buyers in making useof standard search and browsing facilities.Particularly in the case of casual oroccasional buyers and in the case of complexproducts, the gap between a product'sspecifications and the buyer's understanding ofneed can be hard to bridge. An effectivee-commerce catalog must map user needs toproducts that can fulfill them. This paperdescribes an interactive, incremental,case-based, critiquing approach to solving thisproblem. The approach is interactive andincremental, so it does not require that theuser have a completely specified need at thestart. The system is case-based in that itemphasizes products over features orconstraints, and uses case-based reasoningtechniques for its product retrieval. Finally,the approach is based on the critiquing ofpresented examples, each critique redirectingthe search to home in on appropriate products.     

Finding curvilinear features in spatial point patterns: principalcurve clustering with noise

Clustering about principal curves combines parametric modeling of noise with nonparametric modeling of feature shape. This is useful for detecting curvilinear features in spatial point patterns, with or without background noise. Applications include the detection of curvilinear minefields from reconnaissance images, some of the points in which represent false detections, and the detection of seismic faults from earthquake catalogs. Our algorithm for principal curve clustering is in two steps: The first is hierarchical and agglomerative (HPCC) and the second consists of iterative relocation based on the classification EM algorithm (CEM-PCC). HPCC is used to combine potential feature clusters, while CEM-PCC refines the results and deals with background noise. It is important to have a good starting point for the algorithm: This can be found manually or automatically using, for example, nearest neighbor clutter removal or model-based clustering. We choose the number of features and the amount of smoothing simultaneously, using approximate Bayes factors     

差量存储的集中式文件级连续数据保护方法

基于文件的连续数据保护系统可实时捕获单个文件的变化,提供任意时间点的文件恢复。在数据的传输上使用差量算法对文件进行差异传输,存储上使用镜像与差量结合的方式记录文件的变化。该方法有效利用了网络带宽,节约了存储资源。在服务器的工作方式上,采用同步和异步两种方式处理不同命令。对异步命令采用单队列、多处理线程的执行模式,有效提高了短作业响应能力以及多任务并发性能。针对多版本历史文件查找,提出了一种基于索引文件的快速查找方法。此外,对服务的并发执行性能进行了测试与分析。     

An Architecture for Supply Chain Integration and Automation on the Internet

Electronic commerce is happening at a very fast pace and business-to-business ecommerce is taking the lead, a very important part of which is the supply chain integration and automation. There is a high demand for well accepted interoperability standards which need to be fitted together for supply chain integration to meet the business demands such as being able to integrate catalogs from different companies. This will facilitate product comparisons and producing customized catalogs. Given an anchor product anywhere on the supply chain, it should be possible to obtain information on related products that complement or add value to this anchor product. Yet another key issue is the full automation of the supply chain processes. However since a single dominant electronic commerce standard is unlikely, the supply chain integration and automation should be able to accommodate different standards like OBI or OTP. This will make it possible for users to conform to the standards of their choice.Another important fact is that rigid supply chains can co-exist with supply chains formed on the fly where participants can transact business spontaneously since the Web is able to make the information instantly available to all trading partners. Facilitating resource discovery that is discovering information on possible partners and their catalogs on the Internet and transacting business automatically also becomes an important issue.The architecture developed within the scope of this paper addresses these issues. We have used the emerging technologies and standards as the infrastructure of the system proposed; and integrated these to meet the needs of supply chain integration and automation and demonstrated how each of the mentioned functionality can be achieved.     

Soft navigation in electronic product catalogs

Current electronic product catalogs support only hard navigation in the product list. Products or product categories are displayed only if they match a criterion that a user has specified explicitly as a constraint or implicitly by following a navigation link. Hard navigation is problematic if users want to express soft preferences instead of hard constraints. Users will make sub-optimal buying decisions if they mistake soft preferences for hard requirements and focus only on products that match all their preferences. Soft navigation is an alternative means to navigate product catalogs. Users express preferences which are used to evaluate products and display them in such a way that higher-scoring products are more visible than lower-scoring products. This paper presents a product scoring catalog (PSC) that supports soft navigation and allows users to express preferences and rate their importance by following a set of rules. The paper closes by outlining possible extensions to PSC and indicating research issues related to soft navigation product catalogs.     

Mining customer knowledge for electronic catalog marketing

Direct marketing is the use of the telephone and non-personal media to communicate product and organizational information to customers, who then can purchase products via mail, telephone, or the Internet. In contrast, catalog marketing is a type of marketing in which an organization provides a catalog from which customers make selections and place orders by mail or telephone. However, most catalogs for retailing firms are presented to customers in the format of paper catalogs without strategic segmentation design and implementation. In this regard, electronic catalog design and marketing could be a method to integrate the Internet and catalog marketing using market segmentation in order to enhance the effectiveness of direct marketing and sales management in retailing. This paper uses data mining based on association rules from relational database design and implementation for mining customer knowledge. As result, marketing knowledge patterns and rules are extracted for the electronic catalog marketing and sales management of a retailing mall in Taiwan.     

Building and querying e-catalog networks using P2P and data summarisation techniques

One of the critical issues in Web-based e-commerce has been how to efficiently and effectively integrate and query heterogeneous, diverse e-catalogs. We propose an integration framework for building and querying catalogs. Our approach is based on a hybrid of peer-to-peer data sharing paradigm and Web-services architecture. Peers in our system serve as domain-specific data integration mediators. Links between peers are established based on the similarity of the domain they represent. The relationships are used for routing queries among peers. As the number of catalogs involved grow larger, the need for filtering irrelevant data sources will become increasingly high. We apply a summarisation technique to summarise the content of catalogs. The summaries are used to pre-selecting data sources that are relevant to a user query. We use terms e-catalog and catalog interchangeably.