最新刊期
卷 36 ， 期 S1 ， 2008
- 摘要：This paper presents a mechanism of using complex event processing technology to process high level business logic of RFID application.By devising common high level business logic into hierarchal composite event,we automate business logic processing as event detection in RFID middleware.We first address several key issues in event definition and event detection from the perspective of RFID middleware's detection behavior.Then based on colored Petri net,we define Event Flow Detection net System as a formal modeling tool for RFID event detection behavior.We propose constructing rules for event detection model and analy ze detectability of hierarchical composite event based on characteristics of the corresponding net structure.关键词：radio frequency identification(RFID);RFID Middlew are;complex event processing;composite event detection;colored petri net2|1733|0更新时间：2025-07-16
- 摘要：Given requirements upon large scale applications and data security,a kind of architecture of EPCISD isco very system is proposed based on the "centralized indexing" mode.This paper first provides a high-level specification for the system's basic functionality and requirements using the TIOA language.Further,this paper proposes a Pub/Sub based,distributed multi-server architecture for the system to make it load balancing,highly available,and scalable in practical applications,and then formally specifies every component in the distributed system respectively.Finally,this paper makes use of a simulation relation to show that the distributed system satisfies/implements the high level specification.2|1023|0更新时间：2025-07-16
- 摘要：The business component is reused to adapt requirement changes by parametrization and configuration in large scale software for enterprise,but it is developed difficulty because of the large granularity,various parameters and complexity of the structure.This paper presents a pattern based code generation method for the large-grained component to improve the efficiency and correctness of development.A specific formal pattern description language is defined to describe the typical software patterns abstracted from the existing components programmed with multiple programming languages.A code generation algorithm is proposed to transform component models for business semantics into target program based on these patterns.The visualization modeling tool and code generator have been developed to generate automatically business components based on J2EE platform.Result of a case purchase system shows effectiveness of this method.关键词：code generation;pattern;MDA;business component;business component model2|1508|0更新时间：2025-07-16
- 摘要：Face is a typical region of interest of image.The face region hierarchical encryption based on the vision perceptual enhances the pertinence and flexibility of image encryption.A novel face region encryption algorithm of palette image is proposed by involving the palette encrypting and pixels position encrypting.Different keys are used to encrypt the palette and the pixels posi tion,and the hierarchical encryption of palette image is achieved.The first level(lightweight)encryption will keep part of the perceptual information of the original face region and could be used for fast identifying.The second level encryption will eliminate all the perceptual information and achieve a high level security.Further,by combining the public key cryptosystem,a two level secure image transmitting scheme is constructed.关键词：hierarchical encryption;palette image;region of interest;perceptual information2|1158|0更新时间：2025-07-16
- 摘要：As an emerging technology,Image Perceptual Hashing(IPH),is becoming a new hotspot and have broad potential applications.Through extracting the digest of perceptual information of an image and matching based on the digest,IPH supports the identification and authentication of images.Currently,the IPH algorithms in the literature are mainly focused on the image feature extraction,but they do not introduce sufficient perceptual factors.In this paper,several new IPH algorithms based on HVS are proposed.Experiments test and compare the proposed algorithms.The results suggest our methods take more perception information into account and have better performance on robustness and discriminability.关键词：human visual system(HVS);perceptual hashing;image authentication;image quality assessment2|3172|0更新时间：2025-07-16
- 摘要：Adaptive steganography is to embed moderate message bits to gain the optimum trade-off between capacity and imperceptibility.In this paper,an adaptive steganographic method is proposed based on the quantization-based embedding mechanism.The message bits are embedded into triple-pixel differences constructed by the Hilbert curve scanning order through the proposed method.Because the partition codebook used in the quantization-based embedding phase is built up based on the perceptual characteristics,the proposed method can achieve good performance both on capacity and distortion.关键词：steganography;quantization-based emebdding;hilbert filling curve2|1211|0更新时间：2025-07-16
- 摘要：Conventional machine learning methods assume that different learning tasks are isolated,but actually they often relate each other.Transfer learning aims at exploiting such relationships and using historical learning experience to improve the learning on new tasks.Much work has been done regarding transfer learning in many sub-domains of machine learning.This paper surveys the transfer of reinforcement learning.According to theories of cognitive psychology,this paper classifies the transfer technologies of reinforcement learning into behavior transfer and knowledge transfer.This paper analyzes the state-of-the-art technologies and some open problems.关键词：transfer learning;reinforcement learning;knowledge;behavior;cognitive psychology;abstraction;generalization3|5111|0更新时间：2025-07-16
- 摘要：Using multi-scale to measure the similarities between signature sequences is much helpful to obtain high-qualified similarity measures.This paper puts forward a new sequence similarity kernel,called multi-scale sequence spectrum kernel,to measure signature sequence similarity based on shared occurrences of different-scale continuous subsequences.The multi-scale sequence spectrum kernel is optimized through semidefinite program,and used with support vector machine to verify signature directory sequences.The experiments on the benchmark database show the signature verification accuracy has been enhanced,as this approach could automatically optimize the similarity measures with multi-scales,depending on the signature characteristics.关键词：semidefinite program;sequence spectrum kernel;signature verification2|963|0更新时间：2025-07-16
- 摘要：In this paper,we propose a novel weighted clustering algorithm based on Laplacian weight,which can automatically transform the structure information between clustering objects into weights of objects.Because Laplacian weight can indicate the neighborhood structure of original data set,better clustering is achieved.Performed on conventional C-means or fuzzy C-means methods,the proposed Laplacian weighting scheme can effectively improve the clustering performance.In addition,the new algorithm achieves some extra advantages such as robustness to outliers,suitability for class-imbalance data clustering and insensitivity to number of clusters,etc.Experimental results on artificial datasets and UCI machine learning repository validate the effectiveness of the proposed algorithm.关键词：clustering;Laplacian;weighted2|1636|0更新时间：2025-07-16
- 摘要：While the standard SVM has O(l3)time complexity and O(l2)space complexity with the size of the training set,the CVM proposed in 2005 has linear time complexity,and the space complexity of CVM is independent of the training set s size.In this paper we proposed a novel method called RSCVM.We first defined the upper and lower approximation of a binary CVM,then extended the definition to the multi class situation.Hence RSCVM combined the CVM and rough set theory.We also gave some experiment results on several real world data sets that illustrated RSCVM s merit of faster speed and less support-vectors.关键词：rough set;upper approximation;lower approximation;CVM(core vector machine)2|1039|0更新时间：2025-07-16
- 摘要：Semi-supervised learning has received much attention recently.Co-training is a kind of semi-supervised learning method which uses unlabeled data to boost the performance of standard supervised learning algorithms.A novel co-training style algorithm,RASCO(for RAndom Subspace CO-training),is proposed which uses stochastic discrimination theory to extend co-training to multi-view situation.The accuracy and generalizability of RASCO are analyzed.The influences of the parameters of RASCO are discussed.Experiments on UCI data set demonstrate that RASCO is more effective than other co-training style algorithms.关键词：semi-supervised learning;random subspace;stochastic discrimination;co-training;multi-view;random subspace co-training(RASCO)algorithm2|2001|0更新时间：2025-07-16
- 摘要：For studying the rapid and accurate classification method of pile foundation defects,a multi-layer classification method is adopted to improve one -to-one support vector machine(SVM)multiple classifiers structure,build one-to-one two-layer classification model and propose two-layer multiple classification method based on SVM theory.Compared with the BP neural network,the learning and training of two-layer SVM classification is faster and the real-time of the classification processing is better.In addition,it has stronger adaptability in test environment based on small sample and better classification accuracy.The method is fit for analyzing the multi-category classification problem of defects diagnose of pile foundation such as smaller training samples,higher precise and more input and output classification variables,and provides important support to the recognition of pile foundation which had several defects.关键词：pile foundations;support vector machine(SVM);multi2classification;defects diagnose;neural network2|1216|0更新时间：2025-07-16
- 摘要：An encouragement strategy which can achieve global optimal and stable solution simultaneously is presented.In the experiment,the postman problem is selected as the example,and the three strategies which include Shapley value strategy、average share strategy and encouragement strategy are used to compare with.The experimental data show that the Shapley value strategy and average share strategy are very inefficient,and can not ensure the stability of the coalition.However the encouragement strategy is very efficient and can achieve global optimal and stable solution simultaneously.In the end,the superiority of the strategy is analyzed.关键词：multi-agent;coalition;encouragement strategy2|1063|0更新时间：2025-07-16
- 摘要：To evaluate the information security risk,a planning-based information security risk process modeling approach (PISRPMA)was proposed to model the risk process of Information Security.PISRPMA described risk domain and risk problem in planning domain definition language,a modified bifrost engineer reasoned all exploit paths using correlative algorithms and Graphviz Toolkit built a planning exploit graph to model the risk process.关键词：planning;risk process;planning domain definition language2|1276|0更新时间：2025-07-16
- 摘要：For complicated nonlinear and nonstationary time series prediction,the precision will be decreased if a faster processing speed is reached in online SVR.A new time series prediction method is proposed.Local SVR is firstly adopted to make prediction of the time series.Then the forecast residual with the real data stream is calculated.Finally,the residual is estimated with online SVR algorithm to compensate the predicted value with the local SVR.Experimental results showed that the propo sed method outperformed the Online SVR.关键词：time series prediction;online prediction;SVR;residual2|1679|0更新时间：2025-07-16
- 摘要：The movement of different kinds of commodities through the supply chain forms different commodity workflows. The flow trends and main paths of different commodities can be discovered by commodity workflow mining,and thus facilitate the management and optimizing of supply chain processes.On the basis of RFID data sets,an approach to mine commodity workflow is proposed.A Petri net based commodity workflow net is defined,some workflow patterns supported by the commodity workflow net are discussed,a cleaning algorithm and an aggregating algorithm based on RFID data sets are provided,and the mining algorithms of the commodity workflow net are presented.Finally,some essential experiments are conducted.关键词：commodity workflow;commodity workflow mining;RFID data sets;Petri net3|2183|0更新时间：2025-07-16
- 摘要：We introduced individual-based reasoning method in description logics into the publish/subscribe system of RFID information services,and the temporal relations between events were depicted using ontology of time.Semantic-based matching and the matching between composite events and composite subscriptions which have temporal relations were solved by judging the consistency between event assertion set and subscription.Also,a subscribing language and a matching algorithm were proposed.The results of testing show that the efficiency of the matching algorithm is able to satisfy the need of application.关键词：publish/subscribe;Radio Frequency Identification(RFID);description logics;ontology of time2|1008|0更新时间：2025-07-16
- 摘要：In this paper,a hierarchical P2P RFID code resolution network architecture is proposed.Based on modified Kademlia,RFID code resolution service is implemented.By simulating,this paper draws the conclusion that hierarchical P2P RFID code resolution network has such advantages as load balance,scalable,sing le node failure tolerence etc.Finally,future works are given.关键词：Radio Frequency Identification(RFID);code resolution network;hierarchical P2P;Kademlia2|988|0更新时间：2025-07-16
- 摘要：Based Upon the physical model of back scatting signals in Ship Borne Surface Wave Radar(SBSWR),the spectrum broadening equation of the echo signals in SBSWR,which reveals the relations among the movement of radar platform,the radar system parameters,the target physical attributes and target moving characteristics,is theoretically formulated in the terms of Doppler frequency.The influence of a moving radar platform on the high frequency phases of the radar echoes and on the Doppler spectrum structures of the echoes are analyzed in detail.The dynamic geometrical relations embedded between the moving radar platform and the targets with different phy sical dimensions and with different moving states are given.In addition,the expressions of the bro adening spectrum of the signals from the static targets and the moving targets are presented respectively.These essential relations lay a theoretical basis for analy zing the characteristics of sea clutter and of targets,and for resolving the targets from the broad ened first order Bragg lines in SBSWR as well.Finally,the correctness of the mathematic expressions presented in the article is proved by the data gathered from a preliminary field experiment which is conducted to evaluate and demonstrate several aspects of the SBSWR in September 8 1998.关键词：ship-borne SWR;the spectrum broadening;first order Bragg lines;signal resolving in the broadened spectrum2|999|0更新时间：2025-07-16
- 摘要：The machine learning is an important technolo gy of the data analysis.But facing data dynamic fuzziness questions,many learning algorithms faced with stern challenge.This paper mainly aims at the character of dynamic fuzzy of data,gives the kind of dynamic fuzzy machine learning model,the locally linear embedding algorithm and the improved locally linear embedding algorithm.It analyses the algorithms stability,and confirms their effective through the examples.关键词：dynamic fuzzy sets;dynamic fuzzy logic;dynamic fuzzy machine learning2|1063|0更新时间：2025-07-16
- 摘要：MDA based on modeling and transformation brings novel thought to improve software development,but OMG doesn't give the concrete modeling methodology and reference model.This paper proposes an ICEMDA framework of software development for enterprise transforming on demand,including the process-oriented ICE-CIM,object-oriented ICE-PIM,pattern-oriented ICE-PSM and business component based ICE-CODE.The paper presents modeling thinking on every layer of ICEMDA,core concepts and model transformation based on them.The contribution of ICEMDA is helpful to develop and implement the large scale complicated software for enterprises.关键词：model driven architecture(MDA);enterprise model;enterprise software and application(ESA);interoperable;configurable;executable2|2008|0更新时间：2025-07-16
- 摘要：This paper brings forward a hierarchy of documents according to the abstract level of information being carried. Relations between different levels of documents were analyzed,centered on revisable office document and non-revisable document. An attempt to embed structural office document information into non-revisable document and to retrieve the information from it using Tagged PDF was experimented successfully.It shows the possibility and necessity to linkup the two kinds of document and form a consistent standard set.关键词：document format;document processing;UOF;non-revisable document;tagged PDF2|1295|0更新时间：2025-07-16
- 摘要：Aiming at the problem that there is the lack of uniform evaluating environment and standard for content-based video authentication algorithms at present,based on the investigation of video authentication algorithm evaluation standard,in this paper a test platform for video authentication algorithm is presented,which has both the functions of management and algorithm testing.This platform is mainly designed for testing video authentication algorithms in compression domain.Then 20 testing video sequence in different kinds are chosen to compare and evaluate four typical authentication algorithms from many aspects such as robustness and tampering detection capability on this platform.关键词：video authentication;digital watermark;test platform;algorithm evaluation2|1234|0更新时间：2025-07-16
- 摘要：Co-training algorithm is constrained by its assumption that the features can be split into two compatible and independent subsets.However,the assumption is usually violated in real-world application,especially for independence.We discover its real purpose is to find two classifiers with certain accuracy and sufficient diversity to co-train.First,multi-views are created using different term evaluation functions.Second,instead of directly computing the independence between two sub-views,this paper evaluates the independence between two classifiers,trained on them,by using diversity measures indirectly.Thus a pair of classifiers with certain accuracy and greater diversity is selected.The experimental results show two improved algorithms named TV-SC and TV-DC are both outperform another co-training algorithm named Co-Rnd based on random splitting method,and TV-DC outperforms TV-SC.关键词：semi-supervised text categorization;Co2training;features views;diversity measures;labeled documents;unlabeled documents2|1283|0更新时间：2025-07-16
- 摘要：This paper proposes a novel method of 3D solid models rendering by taking advantage of the parallelism and programmability of GPU(graphics processing unit).Rapid mesh generation and simplification is implemented by a programmable pipeline of a modern GPU.Then,under the condition that the topology structure of the mesh is not changed,the mesh is modified to decrease the energy functions value.Experiments show that the method is very valuable for the applications.关键词：general purpose computation on graphics processing unit(GPU);hardware-based rendering;energy function;mesh optimization2|1475|0更新时间：2025-07-16
- 摘要：This thesis analyzes the changes of the word occurrence frequency in models with time,uses the prediction technique in the time series model analysis to obtain the next data and thus estimate the model parameter,and get a new smoothing method,which increases the probability if the word's frequency is increasing,reduces the probability if the word's occurrence frequency is decreasing on temporal dimension.The experimental data show that this new smoothing method is superior to the others.关键词：natural language process;statistical language model;data sparseness;Time series analysis;dynamic distribution2|1031|0更新时间：2025-07-16
- 摘要：Automatic music sty le classification is an important topic in the area of music information retrieval.In this paper, we present a study on automatic classification of Chinese folk songs by regional style which mainly fo cuses on performance of different feature selection method.We did music style classification experiments on 1392 original Chinese folk songs which are collected from 10 different regions.The experiment results show that support vector machine classification performance has a certain advantage among different classifiers without feature selection.Simultaneously,support vector machine get the highest classification accuracy of 83%with active feature selection method,the feature vector dimensions are reduced from 74 to 35 using active feature selection feature selection.Therefore,the selected feature set is more convenient for music analysis.2|1331|0更新时间：2025-07-16
- 摘要：The authors come up with a definition of measuring differentiations between features,and then put forward a method of clustering-based feature selection(Below referred to as CBFS).The time complexity of the method is nearly linear with both the size of dataset and the number of features.Besides,the method is applicable to the selection of features in large dataset.It can particularly handle data with both Nominal and Continuous Features.The results of the experiment on UCI datasets show that the method is effective and practicable.关键词：clustering;differentiation of feature;feature selection2|3945|0更新时间：2025-07-16
- 摘要：A new method for restrain the cro ss terms of time frequency distributions is presented.By introducing a phase adjust function(PAF)based on the WVD,this method can restrain the cross terms induced by the nonlinear character of the signal.And for multi-component signal,the CLEAN technique is adopted to filter out each component by a band pass filter,and the cross terms between the components are restrained.At the same time,this method has a high time frequency convergency,the simulated results demonstrate the validity of it.关键词：cross-terms;PAF;CLEAN technique;time frequency convergency2|1257|0更新时间：2025-07-16
- 摘要：Along with the constant development of the Internet and the ever increasing amount of data,the role of search engines has become increasingly evident.More users rely on search engines to find the information needed.In order to more effectively cluster the search results,thus facilitating the positioning of information among the original unstructured results,a new label based clustering algorithm is introduced in this paper.The key idea is to use the dictionary resource and Dependency Syntax Parsing in NLP to extract the onto logies related to the query.These extracted ontologies will further guide the choosing of centroids in K-means clustering.Furthermore,the various features of K-means algorithm have been fully investigated,and a way of improvement is proposed by using the cluster labels.Experiments show that this algorithm not only yields more effective cluster results but also provides more informative descriptions of the results;meanwhile,the efficiency has also been largely improved.关键词：search results clustering;ontology;label2|1142|0更新时间：2025-07-16
- 摘要：According to the poor quality problem in the region of interest in the reconstructed image which is generated by the implicit ROI(Region of Interest)algorithms in the low bit rate coding in JPEG 2000,an improved method is proposed.The proposed method makes use of the Lagrange interpolation to construct the weight adaptive function,gives a reasonable weight for the ROI code block according to the proportion of the number of ROI wavelet coefficients and the target code ration the ROI code block.While ensuring that more ROI information is encoded with higher priority,the background wavelet effects are reduced.Therefore the data length of the ROI wavelet coefficients in the data stream is enhanced.The experimental results show that the recon structed ROI image quality of the proposed algorithm is much better than the implicit ROI algorithm.关键词：region-of-interest image coding;Implicit ROI;JPEG20002|916|0更新时间：2025-07-16
- 摘要：Determinating the search direction too early will make motion estimation fall into a local optimum value in H.264.An integer pixel search method based on prediction for motion estimation is proposed in this paper.Experimental results show that it can improve search precision and decrease computation time.Intra-frame prediction is another important issue in video coding;we present a fast prediction method based on template-macro-block.Results show that it speeds up coding rate about 80% while not lower image quality.关键词：motion estimation;intra prediction;H.2642|1432|0更新时间：2025-07-16
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0