基于NB的双级分类模型在邮件过滤中的研究

使用朴素的贝叶斯（NB）分类模型对邮件进行分类，是目前基于内容的垃圾邮件过滤方法的研究热点。朴素的贝叶斯在参数之间联系不强的时候分类效果简单而有效。但是朴素的贝叶斯分类模型中对特征参数的条件独立假设无法表达参数之间在语义上的关系，影响分类性能。在朴素的贝叶斯分类模型的基础上，我们提出了一种双级贝叶斯分类模型（DLB，Double Level Bayes），既考虑到了参数之间的影响又保留了朴素的贝叶斯分类模型的优点。同时时DLB模型与朴素的贝叶斯分类模型的性能进行比较。仿真实验表明，DLB分类模型在垃圾邮件过滤应用中的效果在大部分条件下优于朴素的贝叶斯分类模型。     

Design, analysis, and implementation of an agent driven pull-based distributed video-on-demand system

The problem of employing multiple servers to serve a pool of clients on a network based multimedia service is addressed. We have designed and practically implemented a prototype system employing multiple servers to render a long duration movie to the customers. We have employed a multiple server retrieval strategy proposed in the literature [39] to realize this system. In the system, server coordination, client behavior and service facilities are completely controlled by an Agent based approach in which we have used the recent Jini technology. Several issues, ranging from data retrieval from individual server, behavior of the underlying network infrastructure, to client management and resource (client buffers) management, are considered in this implementation. We describe in detail our experiences in this complete design process of every module in the software architecture, its purpose, and working style. Further, the system is shown to be robust amidst unpredictable failures, i.e., in the event of server crashes. The load balancing capability is built-in as a safe guard measure to assure a continuous presentation. We present a comprehensive discussion on the software architecture to realize this working system and present our experiences. A system comprising a series of Pentium III PCs on a fast Ethernet network is built as a test-bed. Through this prototype, a wider scope of research challenges ahead are highlighted as possible extensions. Bharadwaj Veeravalli Member, IEEE & IEEE-CS, received his BSc in Physics, from Madurai-Kamaraj Uiversity, India in 1987, Master's in Electrical Communication Engineering from Indian Institute of Science, Bangalore, India in 1991 and PhD from Department of Aerospace Engineering, Indian Institute of Science, Bangalore, India in 1994. He did his post-doctoral research in the Department of Computer Science, Concordia University, Montreal, Canada, in 1996. He is currently with the Department of Electrical and Computer Engineering, Computer and Information Engineering (CIE) division, at The National University of Singapore, Singapore, as a tenured Associate Professor. His main stream research interests include, Multiprocessor systems, Cluster/Grid computing, Scheduling in parallel and distributed systems, Bioinformatics & Computational Biology, and Multimedia computing. He is one of the earliest researchers in the field of divisible load theory. He has published over 75 papers in high-quality International Journals and Conferences. He had secured several externally funded projects. He has co-authored three research monographs in the areas of Parallel and Distributed Systems, Distributed Databases, and Multimedia systems, in the years 1996, 2003, and 2005, respectively. He had guest edited a special issue on Cluster/Grid Computing for IJCA, USA journal in 2004. He has been recently invited to contribute to Multimedia Encyclopedia, Kluwer Academic Publishers, 2005. He is currently serving the Editorial Board of IEEE Transactions on Computers, IEEE Transactions on SMC-A and International Journal of Computers & Applications, USA, as an Associate Editor. He had served as a program committee member and as a session chair in several International Conferences. Long Chen received the B.E. degree in Electrical Engineering and M.E. degree in Electrical Engineering from the Northwestern Polytechnic University, P. R. China, in 1998 and 2001, respectively, and the M.E. degree in Computer Engineering from the National University of Singapore, Singapore, in 2004. He is currently a Ph.D. candidate at the Department of Electrical and Computer Engineering, the University of Delaware, United States. His research interests include multimedia systems, distributed system, network security, and computer architecture.     

Analysis of thinning algorithms using mathematical morphology

A precise definition of digital skeletons and a mathematical framework for the analysis of a class of thinning algorithms, based on morphological set transformation, are presented. A particular thinning algorithm (algorithm A) is used as an example in the analysis. Precise definitions and analyses associated with the thinning process are presented, including the proof of convergence, the condition for one-pixel-thick skeletons, and the connectedness of skeletons. In addition, a necessary and sufficient condition for the thinning process in general is derived, and an algorithm (algorithm B) based on this condition is developed. Experimental results are used to compare the two thinning algorithms, and issues involving noise immunity and skeletal bias are addressed     

Answering similarity queries in peer-to-peer networks

A variety of peer-to-peer (P2P) systems for sharing digital information are currently available and most of them perform searching by exact key matching. In this paper we focus on similarity searching and describe FuzzyPeer, a generic broadcast-based P2P system which supports a wide range of fuzzy queries. As a case study we present an image retrieval application implemented on top of FuzzyPeer. Users provide sample images whose sets of features are propagated through the peers. The answer consists of the top-k most similar images within the query horizon. In our system the participation of peers is ad hoc and dynamic, their functionality is symmetric and there is no centralized index.     

An adaptive updating protocol for reducing moving object database workload

In the last decade, spatio-temporal database research focuses on the design of effective and efficient indexing structures in support of location-based queries such as predictive range queries and nearest neighbor queries. While a variety of indexing techniques have been proposed to accelerate the processing of updates and queries, not much attention has been paid to the updating protocol, which is another important factor affecting the system performance. In this paper, we propose a generic and adaptive updating protocol for moving object databases with less number of updates between objects and the database server, thereby reducing the overall workload of the system. In contrast to the approach adopted by most conventional moving object database systems where the exact locations and velocities last disclosed are used to predict their motions, we propose the concept of Spatio-temporal safe region to approximate possible future locations. Spatio-temporal safe regions provide larger space of tolerance for moving objects, freeing them from location and velocity updates as long as the errors remain predictable in the database. To answer predictive queries accurately, the server is allowed to probe the latest status of objects when their safe regions are inadequate in returning the exact query results. Spatio-temporal safe regions are calculated and optimized by the database server with two contradictory objectives: reducing update workload while guaranteeing query accuracy and efficiency. To achieve this, we propose a cost model that estimates the composition of active and passive updates based on historical motion records and query distribution. More system performance improvements can be obtained by cutting more updates from the clients, when the users of system are comfortable with incomplete but accuracy bounded query results. We have conducted extensive experiments to evaluate our proposal on a variety of popular indexing structures. The results confirm the viability, robustness, accuracy and efficiency of our proposed protocol.     

Planning Intelligent Responses in a Natural Language System

Intelligent help systems cannot merely respond passively to the user'scommands and queries. They need to be able to volunteer information,correct user misconceptions, and reject unethical requests when appropriate.In order to do these things, a system must be designed as an intelligentagent. That is, a system needs to have its own goals and then plan forthese goals. A system which did not have its own goals would never refuseto help users perform unethical actions.Such an intelligent agent has been implemented in the UCEgo component of UC(Wilensky et al. 1984; Wilensky et al. 1988) (UNIX Consultant), a natural languagesystem that helps the user solve problems in using the UNIX operatingsystem. UCEgo provides UC with its own goals and plans. By adoptingdifferent goals in different situations, UCEgo creates and executesdifferent plans, enabling it to interact appropriately with the user.UCEgo adopts goals when it notices that the user either lacks necessaryknowledge, or has incorrect beliefs. In these cases, UCEgo plans tovolunteer information or correct the user's misconception as appropriate.These plans are pre-stored skeletal plans that are indexed under the types ofsituations in which they are typically useful. Plan suggestion situationsinclude the goal which the plan is used to achieve, the preconditions of theplan, and appropriateness conditions for the plan. Indexing plans bysituations improves efficiency and allows UC to respond appropriately to theuser in real time.Detecting situations in which a plan should be suggested or a goal adoptedis implemented using if-detected daemons. These daemons provide asingle mechanism which can be used both for detecting goals and suggestingplans. Different methodologies for the efficient implementation ofif-detected daemons are discussed.     

基于改进叶素动量理论的水平轴风电机组气动性能计算

综合考虑到风剪切、塔影效应、三维旋转效应的影响,本文对传统的叶素动量理论（bladeelementmomentum,BEM）进行改进,建立风电机组气动性能计算模型,基于该模型编制计算程序,以商用1．5MW风电机组为计算实例,计算出其在不同的风速、转速和桨距角配置下的轴向和切向气动荷载分布,以及推力、功率和风能利用系数,与传统BEM模型及风电设计分析软件FOCUS5计算值对比,验证了该模型的正确性和优越性。     

A Novel Character Segmentation Method for Text Images Captured by Cameras

Due to the rapid development of mobile devices equipped with cameras, instant translation of any text seen in any context is possible. Mobile devices can serve as a translation tool by recognizing the texts presented in the captured scenes. Images captured by cameras will embed more external or unwanted effects which need not to be considered in traditional optical character recognition (OCR). In this paper, we segment a text image captured by mobile devices into individual single characters to facilitate OCR kernel processing. Before proceeding with character segmentation, text detection and text line construction need to be performed in advance. A novel character segmentation method which integrates touched character filters is employed on text images captured by cameras. In addition, periphery features are extracted from the segmented images of touched characters and fed as inputs to support vector machines to calculate the confident values. In our experiment, the accuracy rate of the proposed character segmentation system is 94.90%, which demonstrates the effectiveness of the proposed method.     

Modeling study of the impact of surface roughness on silicon and Germanium UTB MOSFETs

We outlined a simple model to account for the surface roughness (SR)-induced enhanced threshold voltage (V/sub TH/) shifts that were recently observed in ultrathin-body MOSFETs fabricated on <100> Si surface. The phenomena of enhanced V/sub TH/ shifts can be modeled by accounting for the fluctuation of quantization energy in the ultrathin body (UTB) MOSFETs due to SR up to a second-order approximation. Our model is then used to examine the enhanced V/sub TH/ shift phenomena in other novel surface orientations for Si and Ge and its impact on gate workfunction design. We also performed a calculation of the SR-limited hole mobility (/spl mu//sub H,SR/) of p-MOSFETs with an ultrathin Si and Ge active layer thickness, T/sub Body/<10 nm. Calculation of the electronic band structures is done within the effective mass framework via the Luttinger Kohn Hamiltonian, and the mobility is calculated using an isotropic approximation for the relaxation time calculation, while retaining the full anisotropy of the valence subband structure. For both Si and Ge, the dependence of /spl mu//sub H,SR/ on the surface orientation, channel orientation, and T/sub Body/ are explored. It was found that a <110> surface yields the highest /spl mu//sub H,SR/. The increasing quantization mass m/sub z/ for <110> surface renders its /spl mu//sub H,SR/ less susceptible with the decrease of T/sub Body/. In contrast, <100> surface exhibits smallest /spl mu//sub H,SR/ due to its smallest m/sub z/. The SR parameters, i.e. autocorrelation length (L) and root-mean-square (/spl Delta//sub rms/) used in this paper is obtained from the available experimental result of Si<100> UTB MOSFETs, by adjusting these SR parameters to obtain a theoretical fit with experimental data on SR-limited mobility and V/sub TH/ shifts. This set of SR parameters is then employed for all orientations of both Si and Ge devices.     

Effect of elastic recovery on the electrical contact resistance in anisotropic conductive adhesive assemblies

The successful design of anisotropic conductive adhesive (ACA) assemblies depends mainly on the accurate prediction of their electrical contact resistance. Among the parameters that influence this resistance, the bonding force used to compress the conductive particles against the conductive tracks during the assembly process is very important. This paper investigates how the contact resistance changes as the bonding force is removed at the end of the assembly process when the epoxy resin used to bond the surfaces has cured. The final contact resistance is determined by examining, through theoretical, experimental and numerical analyzes, the evolution of the residual stress as the elastic recovery of the compressed conductive particles and tracks takes place when the bonding force is removed. An iterative algorithm derived from methods found in fracture mechanics analysis is used to determine the relationship between the contact resistance, the adhesive strength and the stiffness of the cured resin. It is shown that smaller values of adhesive strength yield higher contact resistance values; and similarly, smaller values of modulus of elasticity of the resin lead to higher contact resistance values.