Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
The Innovative Application of Multiple Correlation Plane

The Innovative Application of Multiple Correlation Plane

Ratings: (0)|Views: 25 |Likes:
Published by ijcsis
Presentation data with column graph and line graph is a well-known technique used in data explanation to compare and show direction that users can easily understand. However, the techniques has limitations on the data describing complex with multiple relations, that is, if the data contains diverse relationships and many variables, the efficiency of the presentation will decrease. In this paper, the mathematical method for multi relations based on Radar graph is proposed. The position of information approaches on the correlation plane referred to the distribution of content and the deep specific content. However, the proposed method analyzes the multi variants data by plotting in the correlation plane, and compared with the base line system. The result shows that the performance is higher than other methods in term of accuracy, time and features.
Presentation data with column graph and line graph is a well-known technique used in data explanation to compare and show direction that users can easily understand. However, the techniques has limitations on the data describing complex with multiple relations, that is, if the data contains diverse relationships and many variables, the efficiency of the presentation will decrease. In this paper, the mathematical method for multi relations based on Radar graph is proposed. The position of information approaches on the correlation plane referred to the distribution of content and the deep specific content. However, the proposed method analyzes the multi variants data by plotting in the correlation plane, and compared with the base line system. The result shows that the performance is higher than other methods in term of accuracy, time and features.

More info:

Published by: ijcsis on Jan 20, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

01/20/2011

pdf

text

original

 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 9, December 2010
The Innovative Application of MultipleCorrelation plane
Julaluk Watthananon
Faculty of Information Technology,
King Mongkut’s University of 
Technology North Bangkok,Bangkok, Thailandwatthananon@hotmail.com
Sageemas Na Wichian
College of Industrial Technology,
King Mongkut’s University of 
Technology North Bangkok,Bangkok, Thailandsgm@kmutnb.ac.th
Anirach Mingkhwan
Faculty of Industrial and Technology
Management, King Mongkut’s University
of Technology North Bangkok,Bangkok, Thailandanirach@ieee.org 
 Abstract
 — 
Presentation data with column graph and line graph isa well-known technique used in data explanation to compare andshow
 
direction that users can easily understand. However, thetechniques has limitations on the data describing complex withmultiple relations, that is, if the data contains diverserelationships and many variables, the efficiency of thepresentation will decrease. In this paper, the mathematicalmethod for multi relations based on Radar graph is proposed.The position of information approaches on the correlation planereferred to the distribution of content and the deep specificcontent. However, the proposed method analyzes the multivariants data by plotting in the correlation plane, and comparedwith the base line system. The result shows that the performanceis higher than other methods in term of accuracy, time andfeatures.
 Keywords-Correlation plane; correlation boundary; correlation plot; Star plot; Radar graph
I.
 
I
NTRODUCTION
 In statistics, bar graph and line graph are common types of graphs employed to explain data analyses, to comparedirections and to represent a set of qualitative data withcorrelation between two variables [1]. Nonetheless,comparative analyses of more than two qualitative variablesand multiple correlations have been increasingly implementedin many fields of work, namely weather conditions, contextconsistency of documents, etc. It is important to have a properform of data presentation that can effectively send messagesacross to readers. One of the commonly used forms of datapresentation is a radar chart that can represent data withcorrelation of over two variables in an effective manner due toits continuity and its ability to clearly compare many aspectsof data plot correlations [2]. However, there are a number of limitations in presenting a larger amount of data with multiplecorrelations. Representatives of those relations need to besought so as to determine appropriate data positions.Generally, there are three methods of selectingrepresentatives of data values with correlation of multiplevariables. The three methods are as follows:
1)
 
Selecting the highest value
: classifying quantitative dataof each variable, and then selecting the most quantitiesvariables, for instance, in order to classify books categories[3], librarians will normally do on the essence of the books.Disadvantage of this method is other contents relating to othertopics are decreased in the importance and deleted.
2)
 
Selecting from the mean
: By this method a value datarepresentative from the mean or neutral value calculating froman outcome of added data divided by data amount. Thismethod is usually employed in research to selecting variablesrepresentatives. However, it is not suitable for selecting datawith multiple correlations because accurate data cannot beidentified clearly.
3) Calculating combined results of directions
: this is ahighly successful technique commonly used with data withmultiple variables [4], [5], [6]. A mathematic process isemployed to acquire relation between rectangular and polar
coordinates on a radar chart and proper coordinates’ positions
resulted from calculations of directions and distances of thoserelations. The authors name these plots
 data correlation plots
.They are on correlation plane of connected lines and willconfine the area, create an
 n
axis and divide the plane withinpolar coordinates. The plane in this research is referred to as
 the correlation plane
. The intersection of 
 n
axis is called theorigin. Intersection of 
 n
axes will divide the plane into
 n
parts.Each part is called a
 correlation boundary
, details of whichare elaborated in Section 3.Hence, the authors have developed a concept of applyingthe method of calculating combined results of directions topresent results in the correlation form as above mentioneddefinition. Furthermore, efficiency of presentation of implementing methods, directions and depth levels of thecorrelation to data with multiple variables was analyzed.The rest of this paper is organized as follows: In thesection 2 we provide a review of related works about stargraph, polar coordinates, distance between plots, Deweydecimal classification and Dewey decimal classification
 – 
 Multiple relations. Section 3, 4 and 5 present the definition of correlation such as: correlation plot, correlation plane and
61http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 9, December 2010
correlation boundary, the concept of application and theexperiments with the discussion on the results respectively.Section 6 is the conclusion.II.
 
A
 
R
EVIEW OF
R
ELATED
W
ORKS
 Many researchers have studied and designed methods of presentation from the information retrieval format that allowsusers to access and easily understand with the visualization,
such as in research Texas State Auditor’s [1] presented how to
use graphs representations of the relationships between two ormore variables and the issues of interest. Yeh [2] presentedstar chart showing the target numeric variable amongcategories. The results showed the GRADAR procedureproviding a new looks to the clinical data and helped checkingthe data, tables and clinical reports. Wang et al. [4], [5]proposed a new graphical representation for multi-dimensionaldata in multivariate. The experimental results showed theeffectiveness of accurate classification. Klippel et al. [8]proposed that the best visual representations for a data setpresented are: how to assign variables to rays and to add colorto rays of a star plot graph. The results shown that the star plotgraphs were meaningful; the represented data and star plotenhanced color had positive effects on the processing speed.Peng et al. [9] presented a new method for shaperepresentation by converting the CSS descriptor circularvector map and defining two histograms in polar coordinatesystem. The advantages of their proposed are simplicity,execution speed and efficiency of well in clustering the shapeimages. Sukmar et al. [12] presented the construction of a newpolygonal interpolant that was based on the concept of naturalneighbors. They used technique to adapt the aboveconstruction on polygonal elements to quad tree meshes toobtain C
o
(
) admissible approximations along edges with
“hanging nodes.” Mohseni et al. [13] presented a method for 
treating the coordinate singularity whereby singularcoordinates were redefined. Thus, the results showed the newpole treatment giving spectral convergence and more accuratefor all. Demsar et al. [14] presented a new method for
visualization “FreeViz”. The results showed that the FreeViz
was very fast and can presented high quality with clear classseparation.From the researches above, the most effective technique topresent data was a compute of the relationships and presenteda new method for intelligent visualization [4], [5], [9], [12],[14], [15], [16], [17] of data sets. In this paper, we also appliedthe star graph and polar coordinates to improve theclassification correlation and presented the position of data.Since a normal plane cannot explain correlations of thatcalculated position as a result of the starting point originatedfrom variables with multiple correlations. Below are theoriesof related works with techniques coming from these diversefields.
 A.
 
Star graph
The star graph (can call radar graph or spider graph) is atechnique used to represent graphical data analysis with allvariables in multivariate data sets. It consists of a different
variable of lines radiating from the center. It means the “datalength” of each variable. The characteristics of radar graph are
polygons under the frame of circle that shows many data in thesame graph, so the principles of creation consist of:
1)
 
 Determination of the axis:
Determination of the axisand number of axis displays data where we define one axis forone data, the first axes is in vertical (x-axis) and thencirculates to the east. In addition, users can define the color[5], [8], weight and name of title.
 2)
 
Plot the value on the axis:
Plot the value on the axisthat starting from the origin (point O) to the circumference byassigning the position (x, y) on each axis.
  B.
 
Polar coordinates
The polar coordinate [4], [5], [6], [9] is a popular methodused to calculate the appropriate location of multi variances, inorder to represent the data referred to multiple relations. Theresearch of Wang et al. [4] shows that, this method can beclassified of data efficiently. In previous works [6], we hadanalyzed and computed the correlation of document contentsby DDC-MR method [3]. It showed that position could refer tothe relationship of multiple variables effectively, so this paperwe used the sum of vector method to represent the multivariances as shown in Figure 1.
Figure 1. Example data with multi variances, where n is the number of variance, r
i
 
is relationships between rectangular and polar Coordinates (r, θ).
 
In Figure 1 we show example data with multi variances, let
ij
 
denote the distance of a
 point P
from the origin and thesymbol
 
O
is the data length. The shade means area in thecomputed appropriate position of multiple variable, let
 
=angle between the radial line for
 P
to
O
and the given line
 
 
= 0”,
a kind of 
 positive axis
for our polar coordinate systemand
 R
is the distance from the point
 P
to the origin. Polarcoordinates are defined in terms of ordinary Cartesiancoordinates by computing and connecting the
n
points
 P
ij
, for
i
= 1,…,n
. It is calculated by using equation as follows:
O P
62http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 9, December 2010
iijiijij
 y xP
    
sincos
(1)
where r ≥ 0
 
0 ≤
< 2
, that every point
P(x
ij
 , y
ij
)
in theordinary
 xy
 – 
 plane
(correlation plane) can be rewrite to
(r,
 
)-
answer which is, is a result of the fact of 
P
lines on thecircumference.From these multiple relations, we called correlation of dataon the coordinates of our point
 P
satisfy the relation
x
ij2
+ y
ij2
 =
ij2
(cos
2
 
i
+ sin
2
 
i
)
 
x
ij2
+ y
ij2
=
ij2
 
(so that, as weindicated, the point
P(x
ij
 , y
ij
)
and
(cos
2
 
i
+ sin
2
 
i
) = 1
) is on acircle of radius
centered at
O
). So, we can find
 
by solvingthe equation as:
,arctantan
    
ijijiijiji
 x y x y
    
(2)where
 
in the interval 0 ≤
< 2
, let
arctangent 
denotingthe function by arctan we see that:
            
,232arctan,22arctanarctan
           
if  x yif  x y x y
ijiji
(3)with the interpretation that
 
= ±
 
 /2
corresponds to pointson the real
 y
 – 
axis
and
 
= 0
corresponds to points on the real
 x
 – 
axis
, that we called
correlation plot 
.
C.
 
The distance between plots
We can use the theoretical
Pythagorus
1
 
method to computethe distance between points in the plane in order to find thedistance
. In Data mining we call
Centroid 
[7] to calculateusing equation as follows:
niii
vcn
1
1
(4)where
is the centroid or the correlation plot
(x
Coordinate
 , y
Coordinate
),
i
 
is coordinates in the circumference
(x
i
 , y
i
),
and
|C - V 
i
|
 
is the distance between plots with the coordinates of 
i
 in the circumference, we see that:
22
icoordinateicoordinatei
y y x xvc
(5)
 D.
 
 Dewey decimal classification
Dewey decimal classification was developed by MelvilDewey in 1876. It is widely used in the library. Besides, thereare many kinds of the books which unlimited of any field.That is the system used in more than 200,000 libraries in 135countries. The Dewey decimal system divides the knowledgeinto 10 classes, in each class it is divided into 10 sub-classesand in each sub-class it is divided into 10 divisionsaccordingly. By using numbers as symbols with the purpose of easy to remember, it is popular to use with more than 30languages translation around the world.
 E.
 
 Dewey decimal classification
 – 
Multiple relaitons
Dewey decimal classification
 – 
Multiple Relations or wecall
DDC-MR
. It is a technical analysis classification multiplerelations which was developed between Search engine andDewey decimals classification. It focuses on the analysis of proportion in the content [3]. By using the library standardclassification schemes, one keyword will be able to classify asdeep as 4 levels which assigns number for notation [6], [7].This scheme refers to DDC that does divide human knowledgeinto 10 classes in the first level, 100 subclasses in the secondlevel, 1000 divisions in the third level and the last level or leaf node contains more than 10000 sections.III.
 
D
EFINITION TO
C
ORRELATION
 Our study of implementing methods is to study of correlation deformation connected by related radar graphs, andsubsequently replaced by polar coordinates. One main concernof the study of implementing methods is to consider theshapes, quantities of content correlations, distances,correlation positions and directions of determined coordinates.Thus, in this research, the authors provide definitions for thepurpose of comparing correlations before and afterdeformation and identifying advantages and implementingmethods. For instance, a document pertaining to manysciences, when examined to find out whether it is a suitablerepresentative of documents, has to be adapted so that the plotposition is found and the plot of intrinsic correlation on theplane and boundary is consistent with that correlation. Assuch, a normal plane cannot explain correlations of thatcalculated position because the starting point originates fromvariables with multiple correlations. Below are definitions of keywords.
 A.
 
Correlation plot 
A correlation plot indicates a position of coordinatesderived from a calculation of combined values of everycorrelation so that one position on the same area is identified.The point resulting from that calculation is titled in thisresearch as a
 correlation plot
, which is used to show orrepresent a position of each data set on the correlation planereferring to any correlation with
n
relevant contents.Correlations can be demonstrated in pairs
(r,
 
 )
, where thefirst pair refers to only one plot and represents only one dataset of distance and directional correlations of variables onpolar coordinates. For example, one document containing anumber of related contents is represented by
n
axis (withresults shown in the form of a radar graph), and thencalculated by mutual tension. Consequently, one plot in theform of 
(r,
 )
was acquired as seen in Figure
 
1
.
1
http://en.wikipedia.org/wiki/Pythagorean_theorem
63http://sites.google.com/site/ijcsis/ISSN 1947-5500

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->