In this paper, we present a new image preprocessing method which is based on Wavelet Transform. Before we extract expression texture information, Wavelet transform will be applied to remove partial high frequency info...
详细信息
ISBN:
(纸本)9781479973989
In this paper, we present a new image preprocessing method which is based on Wavelet Transform. Before we extract expression texture information, Wavelet transform will be applied to remove partial high frequency information. And then Gabor Transform will be used to extract expression information. At last we chose the Support Vector Machine to classify facial expressions. Experiments on the Japanese Female Facial Expression (JAFFE) Database illustrate that the Wavelet Transform and Gabor features are effective for facial expression discrimination. Additionally, the experiment results validate the feasibility of this method.
Micro-blog is easily posted and communicated because of its short text ***,its brevity limits the feature *** paper presents a blogger modeling method for Chinese micro-blog sentiment *** paper would adopt a multidime...
详细信息
ISBN:
(纸本)9781467369541
Micro-blog is easily posted and communicated because of its short text ***,its brevity limits the feature *** paper presents a blogger modeling method for Chinese micro-blog sentiment *** paper would adopt a multidimensional strategy to accomplish the feature extraction *** with traditional feature extraction methods and combined with the interactive property of blog,the blog comment will be considered as the performance characteristic of micro-blog emotion *** comes to comments content and the relationship between commentators and micro-bloggers while it extracts the micro blogging text *** last,the paper adopts fuzzy SVM model to reduce the sensation of noise points and outliers on classification *** results demonstrate the effectiveness of the multidimensional feature extraction method on sentiment classification and the classify effect of fuzzy SVM is better than SVM.
With the development of artificial intelligence and pattern recognition, facial expression recognition plays a more and more important role in intelligent human-computer interaction. In this paper, we present a model ...
详细信息
ISBN:
(纸本)9781479973989
With the development of artificial intelligence and pattern recognition, facial expression recognition plays a more and more important role in intelligent human-computer interaction. In this paper, we present a model named K-order emotional intensity model (K-EIM) which is based on K-Means clustering. Different from other related works, the proposed approach can quantify emotional intensity in an unsupervised way. And then the output from K-EIM is encoded. The coding results are used for the dynamic facial expression recognition. The experiment is conducted on Cohn-Kanade facial expression database and the support vector machine classifier is used for facial expression classification. This method achieved a dynamic facial expression recognition accuracy of 88.32% which suggest that the proposed method shows better performance and proves its validity. Moreover, effect of different segments of emotional intensity is also discussed in the paper.
In this paper,Deep Belief Nets(DBN) model and Support Vector Machine(SVM) are used to mine the features hidden in social news,which can influence the emotions of *** feature selection methods for text modeling are use...
详细信息
ISBN:
(纸本)9781467369541
In this paper,Deep Belief Nets(DBN) model and Support Vector Machine(SVM) are used to mine the features hidden in social news,which can influence the emotions of *** feature selection methods for text modeling are used to build the input vectors of DBN,with the purpose of keeping the text information to the greatest *** take advantage of the deep features abstracted by DBN to build social news text *** the same time,three optimal models are used as inputs of SVM to train and classify the social *** get a conclusion that DBN not only reduces the dimension of original features,but also makes the abstracted features with more text information and shows better performance in determining the influence on people' s emotions by social news.
In order to find the overlapping community structure more quickly and accurately in complex network, this paper proposes a Connected Component-Based Distributed method (CCBD) for overlapping community detection. CCBD ...
详细信息
ISBN:
(纸本)9781479926978
In order to find the overlapping community structure more quickly and accurately in complex network, this paper proposes a Connected Component-Based Distributed method (CCBD) for overlapping community detection. CCBD first divides edges set in the network into smaller one. Next, it seeks out connected component using distributed platforms and gives a serials number to them according to certain rules. Then, it classifies these connected components according to the serials number. Task nodes determine whether any two connected components and non-propagating edges can be connected to become a larger community. Finally, CCBD negotiates a new serials name for the new community. Through repeated iterations, we get the community structure in the network. Nodes belong to multiple communities are overlapping nodes. Experiment results show that CCBD has higher time efficiency benefiting from distributed computing. Moreover, the quality of communities detected by CCBD surpasses those found by other algorithms.
This paper presents a DBN (deep belief nets) model and a multi-modality feature extraction method to extend features' dimensionalities of short text for Chinese micro blogging sentiment classification. Besides tra...
详细信息
ISBN:
(纸本)9781479942732
This paper presents a DBN (deep belief nets) model and a multi-modality feature extraction method to extend features' dimensionalities of short text for Chinese micro blogging sentiment classification. Besides traditional features sets for document classification, comments for certain posts are also extracted as part of the micro blogging features according to the relationship between commenters and posters though constructing micro blogging social network as input information. Then, the integration of the above modality features is combined and represented as input vector for DBN. In this paper, a DBN model, which is stacked with several layers of RBM (Restricted Boltzmann Machine), is implemented to initialize the structure of neural network. The RBM layers can take probability distribution samples of original data to learn hidden structures for better feature representation. A Class RBM (Classification RBM) layer, which is stacked on top of several RBM layers, is adapted to achieve the final sentiment classification. The results demonstrate that, with proper structure and parameter, the performance of the proposed deep learning method on sentiment classification is better than state of the art surface learning models such as SVM or NB, which proves that DBN is suitable for short-length document classification with the proposed feature dimensionality extension method.
Emotion recognition at sentence level is one of the fundamental problems of textual emotion understanding. Based on the observation that sentence emotional focus can be expressed by some clauses in this sentence, this...
详细信息
Emotion recognition at sentence level is one of the fundamental problems of textual emotion understanding. Based on the observation that sentence emotional focus can be expressed by some clauses in this sentence, this paper proposes to And the emotional focus for sentence emotion recognition. For the sake of breaking through the problems brought about by depending on emotion lexicons, we first recognize word emotions in a sentence based on Maximum entropy model. And then homogeneous Markov model is built for clause emotion recognition;After that, a strategy based on emotion selection is proposed for a sentence with multiple clauses, and genetic algorithm is used for clause selection by textual feature weighting. The experimental results show that, comparing with the baseline, there are 9.1% and 3.6% improvement respectively for two different evaluations. It is demonstrated that finding emotional focus by clause selection is able to improve the performance of sentence emotion recognition significantly.
Based on the differently implicational idea, the a-universal multiple I restriction method is put forward for general fuzzy reasoning, which contains the a-multiple I restriction method as its specific case. First of ...
详细信息
Based on the differently implicational idea, the a-universal multiple I restriction method is put forward for general fuzzy reasoning, which contains the a-multiple I restriction method as its specific case. First of all, we give the a-universal multiple I restriction principle, which improves the previous restriction principle, and then provide the existing condition of the solutions of the new method. Furthermore, we obtain the optimal solution of the new method for the fuzzy implications with residual pair, the R-implications, as well as some particular fuzzy implications. Finally, it is found that the new method is more reasonable by contrast with the a-multiple I restriction method.
An event structure acts as a denotational semantic model of concurrent systems. Action refinement is an essential operation in the design of concurrent systems. However, there exists an important problem about preserv...
详细信息
An event structure acts as a denotational semantic model of concurrent systems. Action refinement is an essential operation in the design of concurrent systems. However, there exists an important problem about preserving equivalence under action refinement. If two processes are equivalent with each other, we hope that they still can preserve equivalence after action refinement. In linear time equivalence and branching time equivalence spectrum, step equivalences, which include step trace equivalence and step bisimulation equivalence are not preserved under action refinement [17]. In this paper, we define a class of concurrent processes with specific properties and put forward the concept of clustered action transition, which ensures that step equivalences are able to preserve under action refinement.
An event structure acts as a denotational semantic model of concurrent systems. Action refinement is an essential operation in the design of concurrent systems. But there exists an important problem about preserving e...
详细信息
An event structure acts as a denotational semantic model of concurrent systems. Action refinement is an essential operation in the design of concurrent systems. But there exists an important problem about preserving equivalence under action refinement. If two processes are equivalent with each other, we hope that they still can preserve equivalence after action refinement. In linear time equivalence and branching time equivalence spectrum, interleaving equivalences, which include interleaving trace equivalence and interleaving bisimulation equivalence are not preserved under action refinement [9-11, 14, 16, 21]. In this paper, we define a class of concurrent processes with specific properties and put forward the concept of clustered action transition, which ensures that interleaving equivalences are able to preserve under action refinement.
暂无评论