2003 ADEL Working Paper Series

Visualizing Multi-Dimensional Data

Dr. Ping Chen, Dr. Chenyi Hu, Dr. Heloise Lynn, and Yves Simon

Abstract
High dimensional data visualization is very important in data analysts since it gives a direct and natural view of data. In this paper, we propose a method to visualize large amount of high dimensional data in a 3-D space. In our method, we divide the high dimension data into several groups of lower dimensional data first. Then, we use different icons to represent different groups. Initial experiments on a real data set from oil industry have provided us very encouraging results although further improvements are needed.

Read paper ADEL-WP-03-01

Incorporating Data Mining and Computer Graphics for Modeling of Neural Networks

Richard S.Segall,Ph.D.
Arkansas State University,College of Business,
Department of Economics and Decision Sciences

Abstract
This paper first provides a background on the concepts and development of data mining and data warehousing that need to be known by students and educators.This paper then discusses the applications of data mining for the construction of graphical mappings of the sensory space as a two-dimensional neural network grid as well as the traveling salesman problem (TSP)and simulated annealing.Data mining is also used as a tool for the construction of computer graphics as solutions to the TSP and also for the activation of an output neuron for a three-layer feed-forward network that is trained using a Boolean function.Conclusions and future directions of the research are also discussed.

Read paper ADEL-WP-03-02

Performance of Latent Semantic Analysis

Thao Doan, Russell Deaton, and Ning Zhu, Department of Computer Science and Engineering, University of Arkansas, Fayetteville, AR.
Tom Schweiger, Acxiom Corporation, Fayetteville, AR.

Abstract
Latent Semantic Analysis (LSA) is a matching technique capable of recognizing the semantic relationships of data that ordinary techniques such as string matching cannot. This is especially valuable for data integration applications, like those of Acxiom, where data items are usually related by context, rather than in a literal match. Even though it has been shown that LSA is 30% more effective in finding and ranking relevant pieces of information than existing string-by-string matching techniques (Deerwester et al., 1990; Dumais, 1995), the performance of the LSA seems to be affected by the presence of shared words, or "noise", in data. The objective of this research is to study the influence of noise on the LSA performance quantitatively and analytically, which provides understanding for the following researches to develop a noise-filter method used to improve LSA performance. Our research shows that shared terms degrade the performance of LSA for matching queries to documents from the same category, and result in increased misclassification. In addition, share terms change the document that best matches the query.

Read paper ADEL-WP-03-03

Visualization and Ontology of Geo-Spatial Intelligence Information

Yupo Chan, Li Ning, Ningning Wu, John Talburt

Abstract
The research identifies methods, processes, and guidelines for measuring and evaluating the success and value of using visualization in analytic processes. The research seeks to address a shortcoming of existing spatial data-mining techniques. Analysts must quickly view different data in various modalities and genres. We anticipate large volumes of homogeneous data, moderate volumes of highly heterogeneous data, plus data with high dimensionality and high linkages. In order to view the different data, we need to incorporate an understanding about how people perceive visual images.

Read paper ADEL-WP-03-04




	2003 ADEL Working Paper Series Visualizing Multi-Dimensional Data Dr. Ping Chen, Dr. Chenyi Hu, Dr. Heloise Lynn, and Yves Simon Abstract High dimensional data visualization is very important in data analysts since it gives a direct and natural view of data. In this paper, we propose a method to visualize large amount of high dimensional data in a 3-D space. In our method, we divide the high dimension data into several groups of lower dimensional data first. Then, we use different icons to represent different groups. Initial experiments on a real data set from oil industry have provided us very encouraging results although further improvements are needed. Read paper ADEL-WP-03-01 Incorporating Data Mining and Computer Graphics for Modeling of Neural Networks Richard S.Segall,Ph.D. Arkansas State University,College of Business, Department of Economics and Decision Sciences Abstract This paper first provides a background on the concepts and development of data mining and data warehousing that need to be known by students and educators.This paper then discusses the applications of data mining for the construction of graphical mappings of the sensory space as a two-dimensional neural network grid as well as the traveling salesman problem (TSP)and simulated annealing.Data mining is also used as a tool for the construction of computer graphics as solutions to the TSP and also for the activation of an output neuron for a three-layer feed-forward network that is trained using a Boolean function.Conclusions and future directions of the research are also discussed. Read paper ADEL-WP-03-02 Performance of Latent Semantic Analysis Thao Doan, Russell Deaton, and Ning Zhu, Department of Computer Science and Engineering, University of Arkansas, Fayetteville, AR. Tom Schweiger, Acxiom Corporation, Fayetteville, AR. Abstract Latent Semantic Analysis (LSA) is a matching technique capable of recognizing the semantic relationships of data that ordinary techniques such as string matching cannot. This is especially valuable for data integration applications, like those of Acxiom, where data items are usually related by context, rather than in a literal match. Even though it has been shown that LSA is 30% more effective in finding and ranking relevant pieces of information than existing string-by-string matching techniques (Deerwester et al., 1990; Dumais, 1995), the performance of the LSA seems to be affected by the presence of shared words, or "noise", in data. The objective of this research is to study the influence of noise on the LSA performance quantitatively and analytically, which provides understanding for the following researches to develop a noise-filter method used to improve LSA performance. Our research shows that shared terms degrade the performance of LSA for matching queries to documents from the same category, and result in increased misclassification. In addition, share terms change the document that best matches the query. Read paper ADEL-WP-03-03 Visualization and Ontology of Geo-Spatial Intelligence Information Yupo Chan, Li Ning, Ningning Wu, John Talburt Abstract The research identifies methods, processes, and guidelines for measuring and evaluating the success and value of using visualization in analytic processes. The research seeks to address a shortcoming of existing spatial data-mining techniques. Analysts must quickly view different data in various modalities and genres. We anticipate large volumes of homogeneous data, moderate volumes of highly heterogeneous data, plus data with high dimensionality and high linkages. In order to view the different data, we need to incorporate an understanding about how people perceive visual images. Read paper ADEL-WP-03-04 © Copyright 2005 Acxiom Corporation. All rights reserved. Please read our Detailed Online Privacy Statement.