Use Of Category Information Research Articles

A compact discriminant hierarchical clustering approach for action recognition

In order to improve the accuracy of action recognition, a compact discriminant hierarchical clustering approach and an action recognition new framework are respectively proposed. Firstly, on the bases of low-level features 3D Self-Correlation Histogram of Oriented Gradient in Trajectory (3D_SCHOGT) and 3D Self-Correlation Histogram of Oriented Optical Flow in Trajectory (3D_SCHOOFT), the mid-level semantics possessing purity, representativeness and discriminativeness simultaneously are obtained using the proposed compact discriminant hierarchical clustering approach, in which removal of singularities, quantitative evaluations of purity, representativeness and discriminativeness, as well as additive constraint of information entropies for clusters are conducted respectively to assure the better purity, representativeness and discriminativeness. Secondly, by introducing category constraint, a discriminant classification model of Category Constraint Latent Support Vector Machines (CC-LSVM) is proposed, which enhances the discriminative ability of classifier. Finally, to further improve the accuracy of action recognition, a new framework is proposed, which introduces low-level features, mid-level semantics and mid-level semantic self-correlation features into the proposed CC-LSVM classifier in a weighted association way, makes full use of category information of actions, and mines the correlations between multi-semantic features and action categories. Consequently, the action recognition accuracy is improved. The accuracies on Weizmann, KTH, UCF-Sports and YouTube datasets are 100%, 98.83%, 98.67% and 90.73% respectively, which outperform all those in contrastive methods. Experiments demonstrate the effectiveness of proposed compact discriminant hierarchical clustering approach and new framework.

How to Improve Text Summarization and Classification by Mutual Cooperation on an Integrated Framework

Text summarization and classification are core techniques to analyze a huge amount of text data in the big data environment. Moreover, as the need to read texts on smart phones, tablets and television as well as personal computers continues to grow, text summarization and classification techniques become more important and both of them do essential processes for text analysis in many applications.Traditional text summarization and classification techniques have individually been considered as different research fields in this literature. However, we find out that they can help each other as text summarization makes use of category information from text classification and text classification does summary information from text summarization. Therefore, we propose an effective integrated learning framework using both of summary and category information in this paper. In this framework, the feature-weighting method for text summarization utilizes a language model to combine feature distributions in each category and text, and one for text classification does the sentence importance scores estimated from the text summarization.In the experiments, the performances of the integrated framework are better than ones of individual text summarization and classification. In addition, the framework has some advantages of easy implementation and language independence because it is based on only simple statistical approaches and POS tagger.

A New Supervised Laplacian Eigenmap for Expression Recognition

In order to reduce interference of identity information on facial expression recognition, this paper proposes a new supervised Laplacian Eigenmap method. Before dimension reduction, make full use of category information to modify the neighborhood structure of the samples which is in the high dimension space, optimize neighborhood information of the test samples, and then modify the weight matrix, improve facial expression recognition rate. The experiments on JAFFE and Cohn-Kanade expression database show the method proposed is more effective to represent facial expression feature than the Laplacian Eigenmap or supervised laplacianfaces.

Query modeling for entity search based on terms, categories, and examples

Users often search for entities instead of documents, and in this setting, are willing to provide extra input, in addition to a series of query terms, such as category information and example entities. We propose a general probabilistic framework for entity search to evaluate and provide insights in the many ways of using these types of input for query modeling. We focus on the use of category information and show the advantage of a category-based representation over a term-based representation, and also demonstrate the effectiveness of category-based expansion using example entities. Our best performing model shows very competitive performance on the INEX-XER entity ranking and list completion tasks.

When to Cry Over Spilled Milk: Young Children's Use of Category Information to Guide Inferences About Ambiguous Behavior

Three studies (N = 171) examined preschool children's tendency to use category information to make inferences about ambiguous behavior. Children heard stories in which category information about story characters was manipulated and behavioral information was held constant. Participants were asked to evaluate, explain, and determine the significance of the behavior in question. Children tended to be harsher judges of the same ambiguous behaviors when performed by (a) humans as compared to animals, (b) boys compared to girls, and (c) older children compared to younger children. Results suggest that young children hold differentiated notions of the mental states and dispositions that underlie behavior and that these notions vary as a function of category membership. These findings support the conclusion that even young children can hold and use multiple folk psychologies.

Framing interethnic ideology: Effects of multicultural and color-blind perspectives on judgments of groups and individuals.

In 3 experiments, White American college students received a message advocating either a color-blind or a multicultural ideological approach to improving interethnic relations and then made judgments about various ethnic groups and individuals. Relative to a color-blind perspective, the multicultural perspective led to stronger stereotypes, greater accuracy in these stereotypes, and greater use of category information in judgments of individuals. This increase in between-category differentiation occurred both for attributes that favored the in-group and for attributes that favored the out-group and was also paired in some cases with greater overall positivity toward the out-group. The findings lead us to question the implicit assumption driving the majority of social psychological efforts at prejudice reduction: that the categorization process leads to prejudice, and that the relevance of social categories must therefore be de-emphasized.

Self-modeling mixture analysis of categorized pyrolysis mass spectral data with the SIMPLISMA approach

Windig, W., Heckler, C.E., Agblevor, F.A., and Evans, R.J., 1992. Self-modeling mixture analysis of categorized pyrolysis mass spectral data with the SIMPLISMA approach. Chemometrics and Intelligent Laboratory Systems, 14: 195–207. A new self-modeling approach to the resolution of spectral data of mixtures has recently been described. Since the interactive method is not based on principal component analysis and since it expresses all intermediate steps in the form of spectral plots, it is relatively easy to use. The new method is a pure variable method and is called SIMPLISMA (SIMPLe-to-use Interactive Mixture Analysis). This paper demonstrates the use of SIMPLISMA on data obtained from pyrolysis mass spectrometry of biomass feedstocks. Mass spectral data are more difficult to evaluate since noise in discrete spectra is not so easy to detect as in continuous spectra. Furthermore, the extension of SIMPLISMA towards the use of category information (in this case replicate analysis of the samples) in the algorithm is described.

The use of semantic knowledge in Alzheimer's disease: Evidence for impairments of attention

The ability to access and use semantic knowledge was examined in 25 patients with Alzheimer's disease (AD) and 31 healthy controls. The AD patients exhibited intact automatic activation of semantic relationships on a semantic priming task. They also made active use of category information on a recognition memory test. However, the patients did not use semantic information to generate expectations on the semantic priming task and were inconsistent in their use of category information on the memory test. It is suggested that attentional deficits underlie the observed impairments and may in fact help explain the patterns of performance across tasks differing widely in their overt cognitive demands.

Category Use in Abstract Mental Comparisons

A recent model proposes that at least two processes are involved in mental comparisons; use of analogue and use of category information. The evidence that category information is used is not strong, however. It comes mainly from studies which had subjects overlearn the objects associated with arbitrary categories in the laboratory, and which used concrete dimensions. In the present study, two experiments were conducted to see if subjects spontaneously use natural categories from an abstract dimension (occupational status) in mental status comparisons. Subjects scaled occupations into status categories before each experiment and then compared the status of within-category and between-category occupation pairs. Within-category status comparisons reliably took longer than between-category ones of the same distance, suggesting that natural status categories were used.

The use of category information in perception.

Three tachistoscopic experiments are reported in which presentation of the target stimulus in a letter/digit categorization task was preceded by a briefly exposed priming stimulus (letter or digit). The primer was subject to backward masking from either the target or a pattern mask, and observers were unaware of its occurrence. With a primer duration of 25 ms, when masking was presumed to be at a central level, performance deteriorated when the two items were from different categories. This inhibition effect was reduced when the characters were physically similar. In contrast, there was little evidence of facilitated processing when primers validly cued targets. At shorter primer durations, when masking is presumed to be peripheral in origin, between-condition differences were less marked. An interpretation in terms of an active model of information processing, with utilization of both categorical and physical information extracted from the primer, is proposed.

The use of category information in a memory-search task.

The use of category information in a memory-search task.