JP5062556B2 - Program for executing image generation method for visualizing database and database display device - Google Patents

Description

  The present invention relates to an image generation method, a program, and a database display device for visualizing a database by displaying the entire data of the database having a hierarchical structure on one screen.

  With the spread of information technology, the information accumulated in computers is rapidly increasing in size, diversifying and becoming more complex. As a means of understanding the contents of such information, research on information visualization technology using computer graphics is progressing. It is said that humans obtain about 90% of information from their eyes in daily life. In other words, it is thought that the human eye originally functions as an important medium for collecting information. From this, it is considered that the information visualization technique is effective as a means for transmitting information from a computer to a human.

  In research of information visualization technology, there is a tendency that the data structure of input information is classified and research is individually progressed for each data structure. As an example, Shneiderman categorizes the data structure targeted by information visualization technology into seven types: one-dimensional data, two-dimensional data, three-dimensional data, n-dimensional data, hierarchical data, link data, and time-series data. Is advocating. Among these classifications, visualization of n-dimensional data (hereinafter referred to as “multivariable data”) and hierarchical data is considered as an important issue in information visualization technology, and there are a great many as described below. Research has been published.

(1) Tree structure Many techniques for visualizing hierarchical data draw a tree structure. Non-patent documents 1 to 3 shown below are known as typical techniques. These tree-structured visualization methods have the purpose of displaying detailed information by human interaction so as to be hidden from the top of the hierarchical structure. However, on the other hand, there is a lot of information interference (overlap) on the screen, which is not suitable for the purpose of viewing the entire information “without operation”.
Lamping J., Rao R., The Hyperbolic Browser: A Focus + context Technique for Visualizing Large Hierarchies, Journal of Visual Languages and Computing, Vol. 7, No. 1, pp. 33-55, 1996. Carriere J., et al., Research Paper: Interacting with Huge Hierarchies beyond Cone Trees, IEEE Information Visualization 95, pp. 74-81, 1995. Koike H., Fractal Views: A Fractal-Based Method for Controlling Information Display, ACM Transactions on Information Systems, Vol. 13, No. 3, pp. 305-323, 1995.

(2) Recursive partitioning of screen space On the other hand, the method based on recursive partitioning of screen space is close to the object of the present invention from the viewpoint of expressing information on a screen so as not to overlap. A typical method is Treemaps that express hierarchical data by recursively dividing the screen space into rectangular areas. A problem with Treemaps is that they often generate elongated rectangular areas on the screen. Many improved methods have already been announced. Among them, the Quantum Treemap that represents individual data elements with icons and thumbnail images of the same size is a particularly useful method.
Johnson B., et al., Tree-Maps: A Space Filling Approach to the Visualization of Hierarchical Information Space, IEEE Visualization '91, pp. 275-282, 1991. Bederson B., Schneiderman B., Ordered and Quantum Treemaps: Making Effective Use of 2D Space to Display Hierarchies, ACM Transactions on Graphics, Vol. 21, No. 4, pp. 833-854, 2002.

“Heiankyo View” is a method of representing individual data elements with icons and thumbnail images of the same size and visualizing them so that they do not overlap each other on the screen, like Quantum Treemap. Both methods have very similar characteristics and are objectively evaluated in the following references. "Heiankyo View" has obtained better results than other methods in each item of calculation time, good shape of rectangular area, and similarity of visualization results of similar data.
Itoh T., Yamaguchi Y., Ikehata Y., Kajinaga Y., Hierarchical Data Visualization Using a Fast Rectangle-Packing Algorithm, IEEE Transactions on Visualization and Computer Graphics, Vol. 10, No. 3, pp. 302-313, 2004 . Ito, Yamaguchi, Koyamada, Improvement of computation time and screen occupancy of hierarchical data visualization method by rectangular nested structure, Transactions of Visualization Information Society, Vol. 26, No. 6, pp. 51-61, 2006. JP 2006-318048 A

The important points common to these methods are
(A) By expressing the hierarchical structure, the position on the screen of each data element constituting the hierarchical data is determined.
(A) Furthermore, each data element must be represented by a very small icon on the screen.
That is the point. In order to represent a hierarchical structure and multiple variables simultaneously, it must be valid even under this constraint.

(3) Representative examples of multivariable data visualization techniques The following are typical examples of multivariable data visualization techniques.

(3.1) A method of drawing vertical lines as many as the number of variables on a two-dimensional Cartesian coordinate system, plotting variable values on each of them, and connecting them in a line graph. Parallel Coordinates is especially known. Since this method displays one data element in a line graph, the data element cannot be represented by a small icon or the like, and it is difficult to represent the hierarchical structure at the same time.
Inselberg A., Dimsdale B., Parallel Coordinates: A Tool For Visualizing Multidimensional Geometry, IEEE Visualization '90, pp. 35-38, 1990.

(3.2) Plot each data element in a 3D Cartesian coordinate system with reference to 3 of the multivariables, followed by another 3D Cartesian coordinate system around each plotted point, A method that repeats the operation of plotting each data element with reference to three variables. Worlds within Worlds is especially known. Although this method can display data elements with small icons or the like, the position on the screen is determined by the values of multivariables, so that it is difficult to simultaneously express the hierarchical structure.
Feiner S., Beshers C., Worlds within Worlds: Metaphors for Exploring n-Dimensional Virtual Worlds, ACM Symposium on User Interface Software and Technology (UIST'90), pp. 76-83, 1990.

(3.3) A method in which multivariable data elements are plotted in a multidimensional space, a principal axis of data is obtained using a technique such as principal component analysis, and each data element is projected onto a two-dimensional plane obtained therefrom. One example is Design Galleries. Although this method can also display data elements with small icons or the like, since the position on the screen is determined by the values of multivariables, it is difficult to simultaneously represent the hierarchical structure.
Marks J., et al., Design Galleries: A General Approach to Setting Parameters for Computer Graphics and Animation, ACM SIGGRAPH '97, pp. 389-400, 1997.

(3.4) Many multivariable data visualization methods based on the idea of selecting two arbitrary variables from among multiple variables and plotting them in a two-dimensional coordinate system have been reported. Known as Scatter Plot. However, since the position on the screen is determined by the multivariable values, it is difficult to simultaneously express the hierarchical structure. Moreover, it is not assumed that all of the multivariables are expressed in one coordinate system.

(3.5) There are many known techniques for expressing multivariables of individual data elements by using small pieces called Glyph and changing their shapes and colors. One example is the technique of Ebart et al.
Ebert DS, et al., Automatic Shape Interpolation for Glyph-based Information Visualization, IEEE Visualization '97, 1997.

  The present invention solves the above-mentioned problems of the prior art and provides a method for visualizing information of hierarchical multivariable data in which individual data elements have multivariables and these data elements constitute a hierarchical structure. The purpose is to do.

The present invention is a method of displaying all the individual data of a database having a hierarchical structure on a single screen, wherein each data in the hierarchical database is displayed as a rectangular or square icon, and In a method for visualizing a database, wherein the icons of each data are arranged in a two-dimensional plane so as to have a recursive nested structure,
A first step of receiving, from the database, input data having a tree structure, wherein leaf nodes of hierarchical data constituting the tree structure have n variables ti (0 ≦ i <n, i is an integer) When,
A second step of dividing the rectangular or square icon corresponding to the leaf node into L vertical and m horizontal (where Lm ≧ n) grid-like small regions;
And a third step of assigning different colors to each of the lattice-like small regions obtained by the division.

The third step is, for example,
Giving each of the small areas a different hue H;
Providing saturation S and brightness I to each of the small regions based on any of the following formulas (a) to (c).
(A)
S = f (ti)
I = g (ti)
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1, and the functions f (ti) and g (ti) are 0 ≦ f (ti). ≦ 1, 0 ≦ g (ti) ≦ 1)
For example, the functions f (ti) and g (ti) are like graphs A to D shown in FIG.
(I)
S = Smin + (Smax−Smin) ti 0 ≦ Smin <Smax ≦ 1
I = Imin + (Imax−Imin) ti 0 ≦ Imin <Imax ≦ 1
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1)
(I ') Or,
S = I = a + bt (where a and b are real numbers where a ≧ 0 and b ≧ 0, in principle, a + b = 1, and the variable value ti (0 ≦ i <n) is in the range of 0 ≦ ti ≦ 1. Normalized)
This corresponds to the graph of A in FIG.
(U)
S = Smin + (Smax−Smin) * | 2ti−1 | 0 ≦ Smin <Smax ≦ 1
I = Imin + (Imax−Imin) * | 2ti−1 | 0 ≦ Imin <Imax ≦ 1
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1)
| 2ti-1 | means the absolute value of (2ti-1).
(U ') or
S = I = a + b (1-2ti) (0≤ti≤c)
S = I = a + b (-1 + 2ti) (c≤ti≤1)
(Where a and b are real numbers where a ≧ 0 and b ≧ 0, in principle a + b = 1, c is a real number where 0 <c <1, and the variable value t i (0 ≦ i <n) is 0 ≦ (It shall be normalized to the range of t i ≦ 1)
This corresponds to the graph of FIG.

  Further, among the small regions corresponding to 0 ≦ ti ≦ TH (TH is a predetermined value and 0 ≦ TH ≦ 1), the small region is defined as a first shape, and the small region And the step corresponding to TH ≦ ti ≦ 1 may include a step of setting the small region to a second shape different from the first shape. For example, TH = 0.5.

The present invention is a method of displaying all the individual data of a database having a hierarchical structure on a single screen, wherein each data in the hierarchical database is displayed as a rectangular or square icon, and A program for causing a computer to execute a method of visualizing a database, characterized in that the icons of each data are arranged in a two-dimensional plane so as to have a recursive nested structure,
A first step of receiving, from the database, input data having a tree structure, wherein leaf nodes of hierarchical data constituting the tree structure have n variables ti (0 ≦ i <n, i is an integer) When,
A second step of dividing the rectangular or square icon corresponding to the leaf node into L vertical and m horizontal (where Lm ≧ n) grid-like small regions;
And a third step of assigning different colors to each of the lattice-like small regions obtained by the division.

The present invention is a display device for displaying all the individual data of a database having a hierarchical structure on a single screen, displaying each data in the hierarchical database on the screen as a rectangular or square icon, In the database display device including a processing unit that arranges the icons of the data of the hierarchical structure in a two-dimensional plane so as to have a recursive nested structure,
The processor is
A first step of receiving, from the database, input data having a tree structure, wherein leaf nodes of hierarchical data constituting the tree structure have n variables ti (0 ≦ i <n, i is an integer) When,
A second step of dividing the rectangular or square icon corresponding to the leaf node into L vertical and m horizontal (where Lm ≧ n) grid-like small regions;
And a third step of assigning different colors to each of the lattice-like small regions obtained by the division.

  The present invention is an improved invention of a known hierarchical data visualization method (Patent Document 1, Non-Patent Documents 6 and 7, hereinafter referred to as “Heiankyo View”). “Heiankyo View” is a visualization method that aims to display all the individual data elements that make up the hierarchical structure on a single screen. Whereas “Heiankyo View” displays each data element in the hierarchical data as an icon (a graphic or symbol representing a leaf node), in the present invention, colors corresponding to the number of variables are assigned to each icon. By assigning and adjusting the intensity (brightness and saturation) of this color, the multivariable value of each data element is expressed.

  The expression form of multivariable data adopted by the present invention is also a kind of Glyph in a broad sense. The advantages of large-scale hierarchical multivariable data visualization based on region division and color change as in the present invention and the reason for adopting multivariable expression by simple region division will be described.

  An object of the present invention is to represent the entire picture of large-scale hierarchical data having hundreds or thousands of data elements on one screen. As a result, the area on the screen that can be used to represent individual data elements is reduced. On the other hand, in order to express large-scale data, it is advantageous to generate a region close to a square that can be visually recognized even if the area on the screen is small. In view of this point, the present invention divides each variable assigned to each data element into small areas close to a square, and assigns a hue to each small area.

The program according to the present invention is recorded on a recording medium, for example.
Examples of the medium include EPROM devices, flash memory devices, flexible disks, hard disks, magnetic tapes, magneto-optical disks, CDs (including CD-ROMs and Video-CDs), DVDs (DVD-Videos, DVD-ROMs, DVD-s). RAM), ROM cartridge, RAM memory cartridge with battery backup, flash memory cartridge, nonvolatile RAM cartridge, and the like.

  A medium is a medium in which information (mainly digital data, a program) is recorded by some physical means, and allows a processing device such as a computer or a dedicated processor to perform a predetermined function. is there.

  An object of the embodiment of the present invention is to display a list of data elements belonging to the lowest level of a hierarchical structure in a limited screen space, rather than displaying a parent-child relationship of hierarchical data. Among the above-described hierarchical data visualization methods, the methods of Quantum Treemap (Non-Patent Document 5) and “Heiankyo View” (Patent Document 1, Non-Patent Documents 6 and 7) are described as follows: This is a technique that meets this purpose in that "a graphic that represents a data element with the same size is placed on the screen". "Heiankyo View" has obtained better results than other methods in each item of calculation time, good shape of rectangular area, and similarity of visualization results of similar data.

  FIG. 1 shows an example of hierarchical data visualization using “Heiankyo View”. FIG. 1 shows a “Heiankyo view” in which a database having specific hierarchical data is displayed on one screen. Reference numeral 10 denotes a leaf node (described later) constituting the hierarchical data. The “Heiankyo View” expresses leaf nodes with rectangular or square icons 10 and branch nodes with nested rectangular frames, and arranges the entire hierarchical data on one screen. As shown in FIG. 1, a very large number of icons 10 are displayed on the screen. The vertical and horizontal directions of the icon 10 are, for example, about several to several tens of dots (pixels) and are very small. However, the absolute position on the screen and the nested rectangular frame allow the observer / evaluator to intuitively and instantaneously understand the structure of the database. This is the advantage of “Heiankyo View”. Of course, the contents of the data cannot be directly known from the icon 10, but the observer / evaluator can easily call the contents by clicking the desired icon 10 after understanding the structure of the database. Therefore, there is no problem (the embodiment of the present invention relates to the display processing of the leaf node (icon) 10, and thus the symbols and descriptions of other elements such as branch nodes are omitted).

“Heiankyo View” realizes the visualization result of FIG. 1 by an algorithm that satisfies the following conditions.
-Arrange so that leaf nodes and branch nodes do not interfere. As a result, any data element on the screen can be displayed in a clickable state.
-Arrange so that the occupied area on the screen space is reduced. Thereby, a large amount of information can be expressed in a limited screen space.
“Heiankyo View” will be explained later.

  Next, a method according to an embodiment of the invention (“12 single views”), which is an extended method of “Heiankyo View”, will be described.

  In the “12 single view”, n pieces of leaf nodes represented by small icons (figure or symbol representing a leaf node) 10 in the “Heiankyo view” for the data in which each data element has n variables. By dividing into the above small areas and assigning a unique color to each small area, the values of n variables of each leaf node are expressed. Based on this concept, the “twelve single view” can represent hierarchical multivariable data having hundreds or thousands of data elements on one screen.

  FIG. 2 shows an input data structure assumed by the method according to the embodiment of the invention. In FIG. 2, a square 20 represents a tree-structured branch node, and a circle 21 represents a tree-structured leaf node. A tree structure is an undirected graph that is one type of graph and is simply connected and does not have a cycle. Nodes that do not have children are called leaves. Of the two nodes connected by branches, the one closer to the root is called the parent, the one closer to the leaf is called the child, and the nodes having the same parent are called siblings.

  In the method according to the embodiment of the present invention, as shown in FIG. 2, it is assumed that leaf nodes of hierarchical data constituting an arbitrary tree structure have n variables (under leaf nodes 21). (T0 to tn-1 in the squares of) means that). In the present invention, the value of this variable is expressed as ti (0≤i <n, i is an integer).

Next, the processing procedure of the method according to the embodiment of the invention will be described with reference to FIGS. FIG. 3A shows icons indicating data in a hierarchical database. FIG. 3B shows an explanatory diagram for dividing an icon into small regions and calculating the hue thereof, and FIG. 3C shows an explanatory diagram of a method 1 for calculating the saturation / lightness of the small region. FIG. 3D shows an explanatory diagram of the method 2 for calculating the saturation and lightness of the small area.
FIG. 4 is a processing flowchart according to the saturation / lightness calculation method 1 described below.
FIG. 5 is a process flowchart according to the saturation / lightness calculation method 2 described below.

The icon 10 generated by the “Heiankyo View” is itself a single color. In the method according to the embodiment of the present invention, first, a rectangular or square icon (leaf node) 10 used in the “Heiankyo view” is divided into L vertical and m horizontal grid-like small regions ( (See S1 in FIGS. 3B, 4 and 5). In the embodiment of the invention, L and m are calculated by the following equation (1).
L = [√n] +1
m = [n / √L] +1 (1)
Here, [t] is a symbol representing the maximum integer not exceeding t. The product of L and m calculated by the above formula is always n or more. For this reason, it is possible to express n variable values by assigning each variable to each small region generated by the above method. In the embodiment of the invention, each of n variables is assigned to Lm small regions in order from the upper left. When Lm> n, there are small areas to which no variable is assigned, but these small areas are left blank here.

  In the example of FIG. 3B, the square icon 10 of FIG. 3A is divided into a square shape. The upper left small area 10-0 has t0, the upper right small area 10-1 has t1, the lower left small area 10-2 has t2, the lower right small area 10-3 has t3, Assigned.

  The reason for adopting such a multi-variable expression by simple area division is as follows. The method according to the embodiment of the present invention aims at collective representation of hierarchical data having hundreds or thousands of leaf nodes, so that the area on the screen that can be spent for representing individual leaf nodes is small. Become. On the other hand, in the area division type visualization method for expressing large-scale data, in order to express data elements so that they can be easily recognized even if the area on the screen is small, an area close to a square is used. There has already been a discussion that it is advantageous to generate. Also in the method according to the embodiment of the present invention, with reference to this discussion, each variable assigned to a leaf node can be expressed by a small area close to a square.

  Subsequently, a color is assigned to each small area. Here, an HSI color system that expresses colors with three variables of hue H representing hue, saturation S representing vividness, and lightness I representing brightness is used.

In the method according to the embodiment of the invention, first, n of Lm small regions are selected, and a unique hue is given to these (S2 in FIGS. 4 and 5). Here, the hue H (0 ≦ H <2π) of the i-th (0 ≦ i <n) variable is simply calculated by the following method (S2a-1 in FIG. 4 and S2b-1 in FIG. 5).
H = 2πi / n (2)

  FIG. 3B shows an example in which n = 4. The small region 10-0 has i = 1 and H = 0, and the color is red (R). The small region 10-1 has i = 2 and H = 0.5π, and the color is green (G). The small region 10-2 is i = 3, H = π, and the color is blue (B). The small area 10-3 becomes i = 4, H = 1.5π, and the color becomes purple (P).

  In the method according to the embodiment of the present invention, the saturation S (0 ≦ S ≦ 1) and the lightness I (0 ≦ I ≦ 1) of each small region are subsequently obtained using n variable values of each data element. Is calculated. For example, the following two types of calculation methods are used. In the following description, it is assumed that the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1.

<< Saturation / Brightness Calculation Method 1 >>
Information visualization techniques often require expressions that encourage attention to data elements with large variable values. Therefore, by increasing the saturation and lightness of data elements having a large variable value, calculation is performed so that these data elements are easily noticed.
S = I = a + bt (where a and b are real numbers where a ≧ 0 and b ≧ 0, in principle, a + b = 1, and the variable value ti (0 ≦ i <n) is in the range of 0 ≦ ti ≦ 1. Normalized)
Specifically, a = 0.2, b = 0.8,
S = I = 0.2 + 0.8ti (3)
Thus, the saturation S and the lightness I are calculated (FIG. 3C, S2a-2 in FIG. 4).
According to this method, the greater the variable value, the greater the saturation and brightness.

<< Saturation / Brightness Calculation Method 2 >>
In some cases, the information visualization technique may require an expression that prompts attention to a data element whose variable value is significantly different from the average value or the intermediate value. Here, a calculation method is adopted in which the saturation and lightness of data elements whose variable values are greatly deviated from the intermediate values are increased so that these data elements are easily noticed. Here, the following formula S = I = a + b (1-2ti) (0≤ti≤c)
S = I = a + b (-1 + 2ti) (c≤ti≤1)
(Where a and b are real numbers where a ≧ 0 and b ≧ 0, in principle a + b = 1, c is a real number where 0 <c <1, and the variable value t i (0 ≦ i <n) is 0 ≦ (It shall be normalized to the range of t i ≦ 1)
Is used to calculate the saturation S and the brightness I (S2b-4 in FIG. 5).
Specifically, a = 0.2, b = 0.8, and c = 0.5.
S = I = 0.2 + 0.8 (1-2ti) (0≤ti≤0.5)
S = I = 0.2 + 0.8 (-1 + 2ti) (0.5≤ti≤1)

  However, in this case, the values of S and I in the variable values ti and 1-ti are exactly the same, and the two cannot be visually distinguished. Therefore, it is preferable to visually distinguish the two cases by applying two different shapes in the case of 0 ≦ ti ≦ 0.5 and 0.5 ≦ ti ≦ 1 (2b in FIG. 5). -2, 2b-3). In the example of FIG. 3D, the small areas 10-1 and 10-2 corresponding to 0 ≦ ti ≦ 0.5 are rotated 45 degrees to visually clarify the difference between the two. This is an example, and the present invention is not limited to this. For example, the small areas 10-1 and 10-2 may be circular, triangular, or star-shaped. In short, the first shape (first graphic) is applied when 0 ≦ ti ≦ 0.5, and the second shape (second graphic) is applied when 0.5 ≦ ti ≦ 1, The first shape (first graphic) and the second shape (second graphic) only need to be visually distinguishable.

  According to this method, the saturation / lightness increases as the variable value moves away from the intermediate value. Two types of shapes are used together depending on the magnitude of the intermediate value.

  The embodiment of the present invention can be applied to visualization of a drug database. An example of a useful method for visualizing a drug database is estimation of experimental values for a new drug. Suppose there is a drug whose molecular structure is known and which has not been tested. If a strong correlation can be found between the molecular structure of this drug group and the experimental values, the experimental results for this new drug can be predicted to some extent before conducting the experiment, which in turn can contribute to the reduction of experimental costs. it is conceivable that. From this, the method according to the embodiment of the present invention, in which the correlation between the characteristics of the molecular structure of the drug and the experimental value can be observed both roughly and locally, can be effectively used in various fields where the drug is handled. .

  In the embodiment of the present invention, a method of visualizing hierarchical multivariable data (“twelve single views”) has been shown. This method is an extension of the hierarchical data visualization method "Heiankyo View". The screen layout of the entire data follows the same principle as "Heiankyo View", and icons representing individual data elements are divided into small areas. In addition, by calculating the color of the small area for each variable, multiple variables are expressed simultaneously with the hierarchical structure.

  When the embodiment of the present invention is applied to the visualization of a drug database, it is possible to overlook the correlation between the characteristics of the molecular structure of the drug group and the experimental results. For the drugs extracted from the database, the experimental values to be visualized are extracted, and further classified according to the similarity and commonality in the molecular structure to construct hierarchical multivariable data, which is used as the embodiment of the invention. It can visualize using the method which concerns on.

Outline of “Heiankyo View” “Heiankyo View” is a method of arranging leaf data that constitutes hierarchical data with icons, branch nodes with nested rectangular frames, and arranging the entire hierarchical data on one screen. (See FIG. 1).

The “Heiankyo View” achieves the visualization results shown in FIG. 1 by adopting a screen layout algorithm that satisfies the following conditions.
-Arrange so that leaf nodes and branch nodes do not interfere. As a result, any data element on the screen can be displayed in a clickable state.
-Arrange so that the occupied area on the screen space is reduced. Thereby, a large amount of information can be expressed in a limited screen space.

The outline of the processing of “Heiankyo View” is as follows.
(A) The data is classified into a plurality of hierarchical clusters based on predetermined characteristic information. For example, compounds are classified hierarchically by various algorithms based on molecular descriptors.
(B) The arrangement of icons in a two-dimensional plane is determined so as to have a recursive nested structure based on the classified hierarchical cluster. More specifically, each icon is arranged in a two-dimensional plane so as to have a nested structure in which child nodes among the leaf nodes classified hierarchically are sequentially included in the parent node. In “Heiankyo View”, the leaf nodes in the hierarchical data are first arranged in a grid. Subsequently, a large amount of information can be expressed on a limited screen space by efficiently arranging branch nodes corresponding to the leaf nodes on the screen.
(C) Further, the parent node to which the leaf node belongs is enclosed in a frame, the further parent node to which each node enclosed in the frame belongs is enclosed in a frame, and so on. Thereby, it is expressed which node of which hierarchical structure the leaf node belongs to.

  The present invention is not limited to the above embodiments, and various modifications can be made within the scope of the invention described in the claims, and these are also included in the scope of the present invention. Needless to say.

An example of visualization of hierarchical data by a conventional method is shown. It is explanatory drawing of the data structure used as the input of the process which concerns on embodiment of invention. It is explanatory drawing of the process which concerns on embodiment of invention. It is a processing flowchart concerning an embodiment of the invention. It is another process flowchart which concerns on embodiment of invention. It is a graph which shows the relationship between ti and S or I which concerns on embodiment of invention.

Explanation of symbols

10 Icon (leaf node)
10-0 to 10-3 Small region 20 Branch node 21 Leaf node

Claims (6)

A method of displaying the entire data of a database having a hierarchical structure on a single screen, wherein each data in the hierarchical database is displayed as a rectangular or square icon, and the icon of each data of the hierarchical structure is displayed. A program for causing a computer to execute a method of visualizing a database characterized by being arranged in a two-dimensional plane so as to have a recursive nested structure,
A first step of receiving, from the database, input data having a tree structure, wherein leaf nodes of hierarchical data constituting the tree structure have n variables ti (0 ≦ i <n, i is an integer) When,
A second step of dividing the rectangular or square icon corresponding to the leaf node into L vertical and m horizontal (where Lm ≧ n) grid-like small regions;
A program for executing the third step of assigning different colors to each of the lattice-like small regions obtained by the division. The third step includes
Giving different hues H to each of the small regions;
The program according to claim 1, further comprising the step of giving saturation S and brightness I to each of the small areas based on the following formula.
S = f (ti)
I = g (ti)
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1, and the functions f (ti) and g (ti) are 0 ≦ f (ti). ≦ 1, 0 ≦ g (ti) ≦ 1) The third step includes
Giving different hues H to each of the small regions;
The program according to claim 1, further comprising the step of giving saturation S and brightness I to each of the small areas based on the following formula.
S = Smin + (Smax−Smin) ti 0 ≦ Smin <Smax ≦ 1
I = Imin + (Imax−Imin) ti 0 ≦ Imin <Imax ≦ 1
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1) The third step includes
Giving different hues H to each of the small regions;
The program according to claim 1, further comprising the step of giving saturation S and brightness I to each of the small areas based on the following formula.
S = Smin + (Smax−Smin) * | 2ti−1 | 0 ≦ Smin <Smax ≦ 1
I = Imin + (Imax−Imin) * | 2ti−1 | 0 ≦ Imin <Imax ≦ 1
(However, the variable value ti (0 ≦ i <n) is normalized in the range of 0 ≦ ti ≦ 1) Among the small regions, those corresponding to 0 ≦ ti ≦ TH (TH is a predetermined value and 0 ≦ TH ≦ 1) is defined as the first small region, and among the small regions 5. The method according to claim 2, further comprising a step of setting the small region corresponding to TH ≦ ti ≦ 1 to a second shape different from the first shape. Program . A display device that displays the entire data of a database having a hierarchical structure on a single screen, displaying each data in the hierarchical database on the screen as a rectangular or square icon, and each data of the hierarchical structure In a database display device including a processing unit that arranges the icons in a two-dimensional plane so as to have a recursive nested structure,
The processor is
A first step of receiving, from the database, input data having a tree structure, wherein leaf nodes of hierarchical data constituting the tree structure have n variables ti (0 ≦ i <n, i is an integer) When,
A second step of dividing the rectangular or square icon corresponding to the leaf node into L vertical and m horizontal (where Lm ≧ n) grid-like small regions;
And a third step of assigning different colors to each of the lattice-like small regions obtained by the division.

Images