There is a growing interest in sharing personal opinions on the Web,such as product reviews,economic analysis,political polls,*** research focuses on document-based approaches and documents are represented by ***,due ...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
There is a growing interest in sharing personal opinions on the Web,such as product reviews,economic analysis,political polls,*** research focuses on document-based approaches and documents are represented by ***,due to loss of contextual information,this representation fails to capture the associative information between an opinion and its corresponding ***,several researches focus on sentence-based approaches,which can effectively deal with an attribute-sentiment word pair within one ***,those approaches are unable to process more than one attribute within one *** this paper,we first present an improved sentiment word quantitative method to generate sentiment score for every word in sentiment ***,we propose a novel identification approach of attribute-modifier-sentiment word triple using shallow semantic *** results show the feasibility and effectiveness of our approach.
This paper proposes an optimization algorithm named Join Order Algorithm Using Predefined Optimal Join Order or is called JAPO algorithm to optimize join cost. Optimal join order solutions for all possible join patter...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
This paper proposes an optimization algorithm named Join Order Algorithm Using Predefined Optimal Join Order or is called JAPO algorithm to optimize join cost. Optimal join order solutions for all possible join patterns are predefined and stored in a file using Dynamic Programming with Memorization technique or is called DPM algorithm. JAPO algorithm searches join order solutions from the predefined optimal join orders using hash function instead of traversing all search space. Experiments are conducted and join costs obtained by JAPO algorithm are compared with DPM algorithm and greedy algorithm named GOO. The experimental results show that JAPO algorithm with polynomial time complexity obtains almost 100 percent of optimal join order solutions. DPM algorithm obtains 100 percent of optimal join order solutions with factorial time complexity. GOO algorithm with polynomial time complexity obtains sub-optimal solutions and number of optimal solutions obtained by GOO algorithm decreases when number of relations to be joined is increased.
Recently,great progesses have been made in using discriminative classifiers in object *** specifically,correlation filters(CFs) for visual tracking have been attractive due to t heir competitive performances on both a...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
Recently,great progesses have been made in using discriminative classifiers in object *** specifically,correlation filters(CFs) for visual tracking have been attractive due to t heir competitive performances on both accuracy and *** this paper,the latest and representative approaches of CF b ased trackers are presented in d *** addition,trackers used deep convolutional features are introduced and several famous tracking methods which fine-tune the pretrained deep network are *** evaluate the performances of different trackers,a detailed introduction of the evaluation methodology and the datasets is described,and all introduced trackers are compared based on the mentioned ***,several promising directions as the conclusions are drawn in this paper.
In order to adjust effectively the shape of the tensor product Bézier surfaces and to connect conveniently the surface patches, generalized tensor product Bézier surfaces with local directional control deriv...
详细信息
With the explosive growth of data capacity, large enterprises increasingly rely on business information systems. A highly informational and data-centralized grid company that engages in high-frequency businesses will ...
详细信息
ISBN:
(纸本)9781509034840
With the explosive growth of data capacity, large enterprises increasingly rely on business information systems. A highly informational and data-centralized grid company that engages in high-frequency businesses will face major business failures, incalculable impacts and irreparable damages once data errors occur. The development of an effective high-availability information system has become a hot topic for grid companies. Based on the analysis of pioneer enterprises or businesses (mainly IT enterprises), this paper summarizes the development mode of gradual deployment, and further proposes an extended architecture suitable for large enterprises such as grid companies in order to provide references and suggestions for the gradual deployment of high-availability information systems.
The hybrid cloud model has received considerable attention in recent years. However, it is challenging to schedule computing applications in hybrid clouds because of the various product types and pricing models provid...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
The hybrid cloud model has received considerable attention in recent years. However, it is challenging to schedule computing applications in hybrid clouds because of the various product types and pricing models provided by different cloud providers in conjunction with the complexity of private cloud resource management. In this paper, we focus on a typical scenario to schedule computing requests in hybrid clouds, trying to minimize the costs incurred while keeping the deadline missing of the requests at an acceptable level. We formulate this problem as an online optimization model originally, and by taking advantage of the Lyapunov optimization techniques, we transform it into a one-shot binary linear optimization problem which is much easier to solve. Based on that we develop a hybrid cloud scheduler, and simulation results suggest that our scheduler strikes a good balance between the cost of public clouds and the deadline missing of computing requests. Besides, the proposed scheduler shows a nearly optimal resource utilization rate and a good average scheduling delay.
In recent years, the trend of population aging brings great demand to the service industry for the aged. Meanwhile, the old-age service resources are relatively scarce, which has become an important issue in the devel...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
In recent years, the trend of population aging brings great demand to the service industry for the aged. Meanwhile, the old-age service resources are relatively scarce, which has become an important issue in the development of society all over the world. In order to improve the service quality of the aged, smart care service based on context awareness was proposed in this paper. Firstly, environmental data, physical sign, voice and facial expression data of the aged are collected. And then affective computing and data mining methods were used to these data, so as to find out the demand of the aged and provide them with smart care service. Results show that this method can effectively help the aged.
In visual tracking, developing a robust appearance model is a challenging task due to variations of object appearances such as background clutter, illumination variation and partial occlusion. In existing tracking alg...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
In visual tracking, developing a robust appearance model is a challenging task due to variations of object appearances such as background clutter, illumination variation and partial occlusion. In existing tracking algorithms, a target candidate is represented by linear combinations of target ***, the relationship between a target candidate and the corresponding target templates is nonlinear because of appearance variations. In this paper, we propose a kernelized convex hull based target representation for visual tracking. Namely, a target is represented by a nonlinear combination of target templates in a mapped higher dimensional feature space. The convex hull model can covers the target appearances that do not appear in the target templates. Experimental results demonstrate the robustness and effectiveness of the proposed tracking algorithm against several state-of-the-art tracking algorithms.
Compiling principle is an important course of computer science major,which mainly introduces general principles and basic methods of the construction of compiling programs *** to high demands of the logic analysis abi...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
Compiling principle is an important course of computer science major,which mainly introduces general principles and basic methods of the construction of compiling programs *** to high demands of the logic analysis ability,the course bring abstract and unintelligible experience to many *** it is quite difficult for students to master the main points of this course within the limited class *** on the requirement above,this paper mainly proposed a method of making use of natural language processing in the research and application of compiling process,which utilizes Maximum Probability Word Segmentation algorithm during the process of lexical analysis and syntax analysis,to offer more effective interface between human and *** proposed method can provide students with intuitive and profound knowledge concept in the process of learning how to compile,makes it easier and quicker for students to understand the principle of computer compiling.
Deep learning for face identification-verification application has been proven to be fruitful. Human faces constituted the main information for human identification besides gait, body silhouette, etc. Deep learning fo...
详细信息
ISBN:
(纸本)9781538619797;9781538619780
Deep learning for face identification-verification application has been proven to be fruitful. Human faces constituted the main information for human identification besides gait, body silhouette, etc. Deep learning for forensic face identification could provide quantitative indexes for face similarity measurement between the questioned and the known human faces in cases, which had the advantage of result objectivity without expert experience influences. We studied the deep learning based face representation for forensic verification of human images. Its application strategies and technical limitations were discussed. We proposed a "winner-takeall" strategy in the case of the forensic identification of human images in videos. We expected the theories and techniques for forensic identification of human images in which both qualitative and quantitative analysis methods were included and expert judgment and automatic identification methods were coexisted.
暂无评论