Multi-structure information retrieval method based on transformation invariance

The needs of efficient and flexible information retrieval on multi-structural data stored in database and network are significantly growing. Especially, its flexibility plays one of the key roles to acquire relevant information desired by users in retrieval process. However, most of the existing approaches are dedicated to a single content and data structure respectively, e.g., relational database and natural text. In this work, we propose “Multi-Structure Information Retrieval” (MSIR) approach applicable to various types of contents and data structures by adapting a small part of the approach to data structures. The power of this approach comes from the use of the invariant feature information obtained from byte patterns in the files through some mathematical transformation. The experimental evaluation of the proposed approach for both artificial and real data indicates its high feasibility. Fuminori Adachi: He received his Master of engineering from Osaka University in ’03. He is enrolled in the doctoral course of Osaka University from ’03. His current research interest includes scientific discovery, data mining and machine learning techniques. Takashi Washio, Ph.D.: He received his Ph.D. from Tohoku University in ’88. In ’88, he became a visiting reseacher in Massachusetts Institute of Technology. In ’90, he joined Mitsubishi Research Institute Inc., and is working for Osaka University from ’96. His current research interest includes scientific discovery, data mining and machine learning techniques. Atsushi Fujimoto: He is enrolled in the master cource of Osaka University from ’03. His Current research interest includes correlation analysis, data mining and machine learning techniques. Hiroshi Motoda, Ph.D.: He received his Ph.D. from University of Tokyo in ’72. In ’67, he joined Hitachi Ltd. and has been working for Osaka University since ’96. His current research interest includes scientific discovery, data mining and machine learning. Hidemitsu Hanafusa: He received Master of Engineering from Keio University in ’83. In ’83, he joined The Kansai Electric Power Co. Ins. (KEPCO). He researched on Maintenance Support System at INSS from ’97 to ’02. Now, he is working in KEPCO.     

A CBIR system based on saliency driven local image features and multi orientation texture features

In Content-based Image Retrieval (CBIR), the user provides the query image in which only a selective portion of the image carries the foremost vital information known as the object region of the image. However, the human visual system also focuses on a particular salient region of an image to instinctively understand its semantic meaning. Therefore, the human visual attention technique can be well imposed in the CBIR scheme. Inspired by these facts, we initially utilized the signature saliency map-based approach to decompose the image into its respective main object region (ObR) and non-object region (NObR). ObR possesses most of the vital image information, so block-level normalized singular value decomposition (SVD) has been used to extract salient features of the ObR. In most natural images, NObR plays a significant role in understanding the actual semantic meaning of the image. Accordingly, multi-directional texture features have been extracted from NObR using Gabor filter on different wavelengths. Since the importance of ObR and NObR features are not equal, a new homogeneity-based similarity matching approach has been devised to enhance retrieval accuracy. Finally, we have demonstrated retrieval performances using both the combined and distinct ObR and NObR features on seven standard coral, texture, object, and heterogeneous datasets. The experimental outcomes show that the proposed CBIR system has a promising retrieval efficiency and outperforms various existing systems substantially.     

中英文Internet信息服务系统

针对中英文Ｉｎｔｅｒｎｅｔ信息服务的特点,并根据互连网信息服务业务发展的管理需要,提出了一种实用的中英文Ｉｎｔｅｒｎｅｔ信息服务系统解决方案,阐述了安全管理、计费管理、全文检索等关键技术。     

子母式穿梭车仓储系统配置优化问题研究

针对仓库资源配置涉及因素众多难以优化问题,提出平均吞吐时间和成本集成的配置优化方法,从提升机和穿梭车两个维度分析,结合子母式穿梭车仓储系统运作特点,考虑穿梭车跨层作业,以平均吞吐时间和成本为目标函数,以货架数量、层数、列数、排数以及设备的数量及运行参数为决策变量,构建子母式穿梭车仓储系统配置多目标优化模型。采用非支配排序遗传算法进行求解,得到Pareto最优解。解集为决策变量的集合,根据这些参数,企业考虑实际情况进行仓储区域的规划布置以及设备选型,为推广应用自动化立体仓库提供了有效参考。     

融合Copulas理论和关联规则挖掘的查询扩展

将Copulas理论引入文本特征词关联模式挖掘,提出融合Copulas理论和关联规则挖掘的查询扩展算法.从初检文档集中提取前列n篇文档构建伪相关反馈文档集或用户相关反馈文档集,利用基于Copulas理论的支持度和置信度对相关反馈文档集挖掘含有原查询词项的特征词频繁项集和关联规则模式,从这些规则模式中提取扩展词,实现查询扩展.在NTCIR-5 CLIR中英文本语料上的实验表明,文中算法可有效遏制查询主题漂移和词不匹配问题,改善信息检索性能,提升扩展词质量,减少无效扩展词.     

High-Speed Tools for Global Information Management.II. Specifications and Uses of the Transparent Query Language (TQL) Part IIF. Additional Uses of TQL, Data Structures, Concise Definitions, Conclusions and Acknowledgments

This part, PART IIF [6], concludes the document HIGH-SPEED TOOLS FOR GLOBAL INFORMATION MANAGEMENT. II. Specifications and Uses of the Transparent Query Language (TQL) [1–6]. It describes novel applications of TQL, the key data structures, and contains a dictionary of Transparent Query Language terms. PART IIF references PART IIA [1], PART IIB [2], PART IIC [3], PART IID [4], and PART IIE [5] and contains Conclusions and Acknowledgements.     

海表面盐度卫星微波遥感研究进展

海表面盐度SSS(Sea Surface Salinity)是研究大洋环流和海洋对全球气候影响的重要参量、是决定海水基本性质的重要因素之一。卫星微波遥感可以满足盐度研究过程中大范围、连续观测的需要,国际上统一的认识是选择频率为1.413 GHz的L波段作为盐度遥感的首选波段。目前,国外发展的海面盐度微波遥感反演算法主要有两种:基于海表发射率估算海表盐度的算法和基于贝叶斯定理提出的反演算法。影响盐度反演精度的因素主要有太空辐射、电离层法拉第旋转、大气、海面粗糙度等。其中,海面粗糙度对盐度反演影响很大,海面粗糙度处理模型可以分为3大类:理论算法(间接发射率模型、直接发射率模型)、经验算法、半经验半理论算法(Hollinger 半经验模型、WISE半经验模型、Gabarró模型)。SMOS卫星和Aquarius/SAC\|D卫星的成功发射,将海表面盐度遥感的反演精度控制在0.2 psu以内,通过改进反演算法,有望得到更高的反演精度。     

基于内容的静态语义概念视频检索方法研究

本文针对互联网视频检索技术的发展,阐述了目前主流视频搜索引擎的技术现状,分析了互联网视频检索的关键技术,特别是对于视频特征的提取技术。本文的创新点是提出了一种通用的基于内容的静态语义视频检索方法,该方法可以弥补基于文本视频检索的有关不足,并且在TRECVID的视频概念检索数据的静态语义概念中得到验证,运行稳定。     

基于多处理器数据库系统的查询调度算法

提出一种多重查询调度算法——连续读取特性（CRP）调度算法,该算法应用连续读取特性,使查询的数据可以重组出某些连续关系,从而使之具有连续读取特性,解决了现有算法中有部分系统资源闲置的问题,提高查询速度。仿真实验结果表明,CRP调度算法的系统利用率和系统性能均优于其他算法。     

基于直推式支持向量机的图像检索

为缩小图像的低层特征与高层语义之间的语义鸿沟,基于支持向量机的相关反馈机制受到越来越广泛的关注,但这种方法并没有利用未标记样本的隐含信息.为更好地利用这些信息,提出将直推式支持向量机作为反馈过程中的学习算法.通过分析其所用特征向量的特点,设计一种颜色稀疏特征,并将其与纹理特征结合作为图像描述的特征.实验结果表明该方法较令人满意,同时也说明直推式支持向量机可在文本分类以外的领域取得较好结果.