The scene environment in the cab is complex and changeable, which is easy to be disturbed by the changes of weather, light and other conditions, which brings many difficulties to the driver’s fatigue driving detection. This paper mainly analyzes the specific characteristics of the driving indoor environment, uses some technologies in the field of computer vision and image processing to improve the image quality, uses the Adaboost face detection algorithm based on Haar-like features, to detect and locate the driver’s face in the cockpit, and reduces the face detection area to determine whether the driver looks left and right or bowed. Then, on the basis of detecting the face, the method of gray scale integral projection is adopted to locate the eyes and the mouth, and the Bezier curve is used to fit the contour of the eyes and the mouth, and to judge the driver’s eye opening and the mouth opening by comprehensively analyzing the vertical and horizontal ratio. Later, ocular fatigue analysis was performed by using the PERCLOS(Percentage of Eyelid Closure over the Pupil over Time) criteria and by eyeblink frequency, and mouth fatigue analysis was also performed by the frequency of yawning.

Product detection based on state abstraction technologies in the software product line (SPL) is more complex when compared to a single system. This variability constitutes a new complexity, and the counterexample may be valid for some products but spurious for others. In this paper, we found that spurious products are primarily due to the failure states, which correspond to the spurious counterexamples. The violated products corre-spond to the real counterexamples. Hence, identifying counterexamples is a critical problem in detecting violated products. In our approach, we obtain the violated products through the genuine counterexamples, which have no failure state, to avoid the tedious computation of identifying spurious products dealt with by the existing algorithm. This can be executed in parallel to improve the efficiency further. Experimental results show that our approach performs well, varying with the growth of the system scale. By analyzing counterexamples in the abstract model, we observed that spurious products occur in the failure state. The approach helps in identifying whether a counterexample is spurious or genuine. The approach also helps to check whether a failure state exists in the counterexample. The performance evaluation shows that the proposed approach helps significantly in improving the efficiency of abstraction-based SPL model checking.

Mining online customer reviews is crucial for analyzing product competitiveness. The effective identification of comparative comments is one of the important prerequisites for influencing marketing decision-making. Sometimes, customers publish a number of comparative reviews after they purchased products, especially if they bought similar products before. Mining these comparative reviews requires a sentimental mining method. A comparative sentence in a review can be transformed into a sentiment score of a particular feature of both compared products. The purpose of this paper is to calculate such emotional scores of product features using comparative sentences and to design a clustering method for analyzing the hierarchical relationship between different brands. To ensure better accuracy of the unsupervised algorithm, an improved computing model based on the sentiment dictionary has been used to obtain weighted sentiment scores. Then, these scores were entered into a design structure matrix, which was used for clustering different brands of similar products. Experiments with a sample of Taobao customer reviews showed that the proposed method can analyze comparative relationships more accurately than conventional methods. This developed design structure matrix is not only suitable for engineering applications, but also for product clustering.

Due to the negative impact from spatial correlation, spatially correlated cognitive radio (CR) based devices participating in cooperative spectrum sensing may be harmful to the detection performance. In this paper, we propose an energy-efficient cooperative spectrum sensing scheme based on spatial correlation for cognitive Internet of Things (CIoT). To mitigate the communication overhead and ensure sufficient sensing accuracy, the CR-based devices (CRDs) can be grouped into several clusters. The member nodes undertake cooperative spectrum sensing tasks in turn, and send the local test statistic to their cluster head nearby. Then, by exploiting the spatial correlation of the members, the cluster head combines the sensing results and makes use of likelihood ratio test to obtain the cluster decision. After receiving the decisions from all clusters, the fusion center employs hard fusion scheme to make the final decision about spectrum occupancy. The simulation results show that our scheme not only provides the better sensing performance, but also improve the energy efficiency.

Real-time flow transmission is widely used in various online applications, such as video transmission, social networks, etc. Most of these tasks which require strict deadlines need to be handled by distributed systems, so the deadline guarantee of the whole system can be divided into each node. However, due to insufficient computing power or software errors, some nodes are prone to delay data processing, leading to the loss of computing deadline, resulting in the loss of the cutoff event of the whole system, and causing the data processing delay of the next link. In view of the performance loss caused by this situation, this paper proposes a cache-based processing framework to speed up the forwarding of real-time data flows. By estimating the data processing ability of nodes online, when the processing ability of nodes is out of order, some computing tasks can be mapped to other preparatory nodes online, so as to achieve fast task. Migration ensures the deadline of computing tasks on nodes. By adjusting reasonable system parameters, the processing capacity of nodes in distributed system can be accelerated, so that the processing capacity of the system can be increased dynamically and the data flow forwarding capability of the whole system can be accelerated. The results show that this method can increase the number of data forwarding by 23% and reduce the data loss rate by 31%.

Automatic expression recognition of human faces has been an active research area for decades. In this work, to improve the facial expression recognition effect, a new method based on SSVM algorithm and multi-source texture feature fusion using KECA is proposed. Multi-source texture features are introduced to describe the facial expression, containing GMCL, Gabor feature, and HOG feature. The results indicate that multi-source texture features are conducive to improve the recognition effect and make up for the deficiency of single texture feature in facial expression description. In addition, KECA and SSVM algorithms show better performance than traditional methods in feature extraction and classification. To further verify the effectiveness of the proposed method, three sets of comparative experiments are carried out: PCA+SVM (based on Gabor feature), PCA+SVM (based on GMCL+Gabor+HOG feature), KECA+SVM (based on GMCL+Gabor+HOG feature). The results of JAFFE database indicate that the accuracy of proposed method, equal to 93.04%, is at least 2.61% higher than conventional method. The results of JAFFE database demonstrate the validity of proposed method.

Mass spectrometry (MS) has played a vital role across a broad range of fields and applications in proteomics. The development of high-resolution MS has significantly advanced biology in areas such as protein structure, function, post-translational modification and global protein dynamics. The two most widely used MS ionization techniques in proteomics are electrospray ionization (ESI) and matrix-assisted laser desorption/ionization (MALDI). ESI typically yields multiple charge values for each molecular mass and an isotopic cluster for each nominal mass-to-charge (m/z) value. Although MALDI mass spectra typically contain only singly charged ions, overlapping isotope patterns can be problematic for accurate mass measurement. To overcome these challenges of overlapping isotope patterns associated with complex samples in MS-based proteomics research, deconvolution strategies are being used. This manuscript describes a wide variety of deconvolution strategies, including de-isotoping and de-charging processes, deconvolution of co-eluting isomers or peptides with different sequences in data-dependent acquisition (DDA) and data-independent acquisition (DIA) modes, and data analysis in intact protein mass determination, ion mobility MS, native MS, and hydrogen/deuterium exchange MS. It concludes with a discussion of future prospects in the development of bioinformatics and potential new applications in proteomics.

Data centers provide services for various real-time applications, such as social networks, instance message, which produce a large number of bursty and urgent mice flow. However, the traditional flow queuing model cannot schedule these flow effectively, which leading the performance loss for some important applications. In this paper, a preemptive scheduling in data center for bursty flow (PSBF) was proposed. The scheduling leverage the preemptive scheduling for more critical flow and set the right queue for other data flow to keep the high aggregated throughput. The experiment and analysis show that our scheduling can handle bursty flow better than traditional flow scheduling for data center.

实践表明,传统的教室和机房的电脑以及学校的服务器会随着时间和使用率快速的老化,而更新的成本十分昂贵,本文提出了运用桌面虚拟化技术进行区域基础教育平台的搭建方案.通过成熟的云平台的安全可靠性和桌面虚拟化技术的灵活性、共享性等优势,来解决传统基础教育领域软硬件更新的问题.我们比较了多个著名的桌面虚拟化系统,并最终选择Citrix技术.最后详述了区域基础教育桌面云系统的基本架构设计,应用Citrix技术构建了面向区域基础教育的桌面虚拟化平台.

Slow response and recommended movies inconsistent with the users'requests are key urgent problems in current movie recommendation system. To address this problem, a context-aware movie hybrid recommendation method on Spark platform is proposed. The method takes advantage of Spark, a distributed parallel computing technology, to improve retrieval and calculation speed for mass data, which reduces the response time of the recommendation system. At the same time, it fuses the user's context information and ALS(Alternating Least Squares of collaborative filtering)to a hybrid recommen-dation method, which improves the recommendation accuracy of system. The results show that our method has a better performance than others.

In order to solve the problem of existing spatial cardinal direction relation expression models which have disadvantage to capture imprecision and fuzziness, the fuzzy description logic and fuzzy rules are introduced into the expression of cardinal direction relation. Fuzzy rules based reasoning approach is proposed to strengthen of imprecision and fuzziness of processing and analyzing ability. In that way, the representing of fuzzy and crisp relations between fuzzy regions or crisp regions can be unified in one formal framework. In the implementa-tion, on the one hand, fuzzy rules are used to rewrite the existing spatial reasoning rule. All rules used in this paper are independent of spe-cific issues, that is to say, all the cardinal direction reasoning are using the same set of rules. On the other hand, the existing annotation of OWL2 is used to realize fuzzy ontology. No new label is extended into OWL2 language. Therefore, there is no need to develop a new rea-soning tool or plugin. There is also no need to prove the correctness of our approach. For reasoning about cardinal relations between fuzzy regions, some approximate methods are used. The experiment results show that the recall ratio of the results is 86.4% when two fuzzy re-gions are used.

With the development of mobile computing, a more intelligent location aware access control approach which can not only protect the privacy of location information but also reduce the frequency of access control decision is needed. This paper brought semantic access control approach into the Location Aware Access Control (LAAC). First, we used semantic area to substitute physical location for the authorization decision, reducing the frequency of authorization decisions. The exact physical location of the subject would not be exposed. Secondly, for the different representation of semantic area of different security domains, we used ontology cluster to represent semantic spatial information. We established security domain ontology (SDO) and bridge ontology (BO) between the security domains. And then we used such ontology cluster as knowledge base and added annotation to the physical areas of XACML language. Finally, the model and framework of location aware semantics based access control for mobile computing is proposed. The proposed location aware semantic access control(LASAC) approach provides a better solution for location aware access control in mobile network environment.

Digital image tamper detection technology is a hot topic in recent decades which has widely applications in image processing. Study of key approach of digital image temper detection has important science and application value. A CFA interpolation based image blur tamper detection algorithm is proposed in this paper. First we reconstruct the tamper image by CFA interpolation artifacts. Then we filter the reconstructed image by Wiener filter and calculate the related degree between the tampering image and the image to be detected. Finally we classify the degree of correlation feature and locate the tampered area. Experimental result shows that the algorithm can effectively detect and locate tampering area of the image been operated by blur tampering and it has a good robustness. The algorithm can accurately locate the image of tampering falsification area and the detection accuracy is significantly better than existing algorithms and can be well positioned image tampering area.

The ontology is usually the knowledge infrastructure for semantic Web and some intelligent systems. Ontology technology is one of the key technologies of the semantic Web. However, the performance of large-scale ontology reasoning has become one of the bottlenecks of the semantic Web based applications. We examine the problem of paralleling the reasoning process for ontology. A key challenge in the problem is partitioning the computational workload to minimize duplication of computation and data communicated among computing nodes. A hybrid partitioning based parallel ontology reasoning approach is proposed to address this challenge. Firstly, OWL ontology files were stored into the database in the form of triple table using Oracle 11g semantic technology. Secondly, a hybrid ontology partitioning approach, which combines both tuple partition and rule partition was designed and implemented. Thirdly, hybrid partitioning based ontology reasoning was achieved using parallel computing technology. Experimental results show that the parallel reasoning methods described herein can significantly improve reasoning performance.

Genetic algorithm is easy to fall into local optimal solution. Simulated annealing algorithm may accept nonoptimal solution at a certain probability to jump out of local optimal solution. On the other hand, lack of communication among genes in MapReduce platform based genetic algorithm, the high-performance distributed computing technologies or platforms can further increase the execution efficiency of these traditional genetic algorithms. To this end, we propose a novel Phoenix++ based new genetic algorithm involving mechanism of simulated annealing. Simulated annealing genetic algorithm has two distinctive characteristics. First, it is the synthesis of the conventional genetic algorithm and the simulated annealing algorithm. This characteristic guarantees our proposed algorithm has a higher probability of getting the global optimal solution than traditional genetic algorithms. The other is that our algorithm is a parallel algorithm running on the high-performance parallel platform Phoenix++ instead of a conventional serial genetic algorithm. Phoenix++ implements the MapReduce programming model that processes and generates large data sets with our parallel, distributed algorithm on a cluster. The experiments indicate that the convergence speed of GA algorithm is significantly faster after adding the simulated annealing algorithm on Phoenix++ platform.

Semantic Web services have brought great convenience to service-oriented software development. However, during the semantic Web service composition because the component Web services and licensing issues often require repeated dynamic binding, which greatly affect the efficiency of the service execution. To address this problem, we propose a defeasible policy based a ccess control approach for semantic Web service composition. Firstly, before the semantic service is bound to a component of Web services, static analysis can avoid unnecessary service binding in the semantic Web service composition and execution time. Then we give the access control enforcement process in composition and execution time. Finally, the feasibility of this method has been verified through experiments. Our approach can increase the efficiency and successful rate of semantic Web service compositio n.

Workload hotspot detection is a key component of virtual machine (VM) management in virtualized environment. One of its challenges is how to effectively collect the resource usage of VMs. Also, since data centers usually have hundreds or even thousands of nodes, workload hotspot detection must be able to handle a large amount of monitoring data. In this paper, we address these two challenges. We first present a novel approach to VM memory monitoring. This approach collects memory usage data by walking through the page tables of VMs and by checking the present bit of page table entry. Second, we present a MapReduce-based approach to efficiently analyze a large amount of resource usage data of VMs and nodes. Leveraging the power of parallelism and robustness of MapReduce can significantly accelerate the detection of hotspots. Extensive simulations have been performed to evaluate the proposed approaches. The simulation results show that our approach can achieve effective estimation of memory usage with low overhead and can quickly detect workload hotspots.

More and more people begin to own multiple computing devices. File synchronization technologies are needed to effectively manage data which spans multiple devices. However there is not a general standard for file synchronization system. In this paper we propose a file synchronization model based on cloud storage - SyncCS. We describe the SyncCS architecture and present a novel two- stage file synchronization protocol along with a conflict resolution mechanism. We evaluate the performance of our proposed file synchronization protocol. Our experimental results indicate that the number of operations to be synchronized when using our protocol is relatively smaller than that using a widely used method.