A Novel Stochastic Attributed Relational Graph Matching Based on Relation Vector Space Analysis

Author(s):  
Bo Gun Park ◽  
Kyoung Mu Lee ◽  
Sang Uk Lee
2012 ◽  
Vol 215-216 ◽  
pp. 270-274 ◽  
Author(s):  
Song Qiao Tao ◽  
Wei He

3D CAD model retrieval has received a lot of attentions in the academic community. Most existent methods for 3D model similarity assessment focus on component models instead of assembly models. In this paper, an assembly model similarity assessment method is presented in order to find the similar assemblies model for design reuse. First, assembly model is described as component attributed relational graph. Then, the compatibility matrix between two assemblies is calculated, which serves as the measure of their similarity. Finally, the optimal matching under the measures is calculated using Hungarian Method. Experimental results show that this method is able to support the assembly similarity evaluation.


2010 ◽  
Vol 43 (3) ◽  
pp. 914-928 ◽  
Author(s):  
Duck Hoon Kim ◽  
Il Dong Yun ◽  
Sang Uk Lee

1989 ◽  
Vol 62 (4) ◽  
pp. 823-833 ◽  
Author(s):  
P. M. Di Lorenzo

1. In the study of the neural code for gustation, it has been suggested that the pattern of responsiveness across fibers or units in the neural pathway for taste may provide the basis for identification and discrimination among taste qualities. Two possible mechanisms of comparison between pairs of stimuli were discussed, as follows: 1) a labeled-line code, where one subset of units responds to one stimulus but not the other, and a second subset of units responds in just the opposite fashion; and 2) a frequency code, where all units always respond well to one stimulus and always respond poorly to the other. 2. Conventional analyses of across unit patterns of response in the taste system have employed the Pearson product-moment correlation and/or the neural mass difference as measures of similarity. These measures consider the relative firing rates for a given pair of stimuli (correlation) or the averaged absolute differences in the firing rates (neural mass difference) evoked by two stimuli. Neither of these metrics considers both the absolute and the relative strengths of response to a given pair of stimuli in the comparison of across unit patterns. 3. A new approach to the analysis of across unit patterns of response, called vector space analysis, was described. With this method, the responses to a given stimulus across units are viewed as a vector in n-dimensional space, where n is the number of units in the sample. The length of each vector provides an index of the overall strength of the response to a particular tastant, and the angle between two vectors is a measure of the similarity of the across unit patterns for a given pair of tastants. 4. A neural discrimination (delta, delta) was derived from this approach as a measure of the similarity of two vectors that incorporates information about both the overall magnitude of response and the distribution of responses across units for a given pair of of stimuli. A labeled line index (lambda, lambda) was also proposed to indicate the extent to which the discrimination between two stimuli may be encoded by the responses in two separate subsets of units. 5. Electrophysiological responses to representatives of the four basic taste qualities (salty, sour, sweet, and bitter) were recorded in 47 single units located in the parabrachial nucleus of the pons (PbN) of the rat. Conventional and vector space analyses were applied to the across unit patterns that were recorded from these units. Multidimensional scaling techniques were used to compare the results of each analysis.(ABSTRACT TRUNCATED AT 400 WORDS)


2010 ◽  
Vol 8 (3) ◽  
pp. 31-46 ◽  
Author(s):  
Kwong-Hung Lai ◽  
Howard Leung ◽  
Zhi-Hui Hu ◽  
Jeff K.T. Tang ◽  
Yun Xu

One of the difficulties in learning Chinese characters is distinguishing similar characters. This can cause misunderstanding and miscommunication in daily life. Thus, it is important for students learning the Chinese language to be able to distinguish similar characters and understand their proper usage. In this paper, the authors propose a game style framework to train students to distinguish similar characters. A major component in this framework is the search for similar Chinese characters in the system. From the authors’ prior work, they find the similar characters by the radical information and stroke correspondence determination. This paper improves the stroke correspondence determination by using the attributed relational graph (ARG) matching algorithm that considers both the stroke and spatial relationship during matching. The experimental results show that the new proposed method is more accurate in finding similar Chinese characters. Additionally, the authors have implemented online educational games to train students to distinguish similar Chinese characters and made use of the improved matching method for creating the game content automatically.


Sign in / Sign up

Export Citation Format

Share Document