JP5754676B2 - Search tree drawing apparatus, search tree drawing method and program - Google Patents

Description

  The present invention relates to a method for drawing a tree structure of a search tree.

  In a database or file system, data is managed using a search tree such as B-Tree or R-Tree in order to search for desired data from the accumulated data set at high speed. B-Tree is a search tree that is an index of an attribute value set of one type of attribute, and R-Tree is a search tree that is an index of an attribute value set of multiple types of attributes. There are also improved techniques such as B + -Tree for B-Tree and RD-Tree for R-Tree.

  In the R-Tree, all data is managed for a multi-attribute data set having the same type of attribute, but a UBI-Tree for managing a heterogeneous multi-attribute data set having different types and different numbers of attributes. Recently, a search tree has been devised (Patent Document 1 and Patent Document 2). As a feature, UBI-Tree manages data in a format in which an attribute name and an attribute value are freely expressed by a sequence of “key = value”, which is a u Tuple format. At this time, “key = value” is synonymous with “attribute name = attribute value”. For example, by expressing as “sensor ID = 1, time = 3, temperature = 20”, three attribute names of sensor ID, time, and temperature, and corresponding attribute values are expressed.

  In general, a search tree has a feature of managing attribute names and attribute value ranges included in a data set included in each node at a lower level with respect to a data set to be managed. When there is one type of attribute as in B-Tree, the normal attribute name is omitted, and only the attribute value range included in the data set included in each node is managed.

  On the other hand, as a search tree visualization method for visualizing the tree structure of the search tree, figures representing the nodes of the search tree are arranged on a cone and displayed three-dimensionally, and the figures representing the nodes are connected by straight lines called edges. The Cone Tree (Non-patent Document 1) representing the relationship between nodes, the figure representing each node and data of the search tree is represented by a rectangle, and the rectangle corresponding to each node and data is inside the rectangle corresponding to the parent node. There is a data jewel box (Non-Patent Document 2) that represents the tree structure of the search tree on a two-dimensional plane by arranging in the scrambler.

  In addition, the data jewel box has an attribute value or attribute of one attribute included in the data in a direction perpendicular to the two-dimensional plane in which the graphic exists, with respect to the graphic representing the data or node managed by the search tree. A function that represents a rectangular parallelepiped or a cone having a length corresponding to the size of an attribute value aggregate value of one attribute included in the data set included in the lower level, that is, the maximum value or average value of the attribute value. It is possible to list the attribute values and attribute value aggregate values of one attribute in addition to the tree structure of the search tree by visualizing it using this function together.

JP 2011-170460 A JP 2011-170461 A

George G. Robertson, Jock D. McInlay, Stuart K.M. Card, “Cone Trees: Animated 3D Visualizations of Hierarchical Information,” Proceedings of ACM Conference on Human Factors in Computing Systems (CHI′91). 189-194. Takayuki Ito, Yasumasa Tominaga, Yuko Ikebata, “Data Jewel Box: Graphics Showcase of Large-Scale Hierarchical Data,” Information Processing Society of Japan, Graphics & CAD Study Group, 2001-CG-104, pp. 65-70, 2001.

  In research tree development and search tree understanding / learning, or checking the status of a search tree and the management status of a data set performed by a database administrator, in addition to the tree structure of the search tree, the data set that each node contains below It is desirable to have an overview of the attribute value range for the attributes included in. In particular, in the case of managing a heterogeneous multi-attribute data set such as the UBI-Tree, the number of attribute types included in a data set included in each node is different for each node. In addition to the above, it is important to visualize so that it is possible to overlook which attribute is included in the data set included in each node and what attribute value range each node has.

  However, in the conventional visualization method, there is only a method in which the type of information that can be drawn at a time is limited or a method in which the area in which information can be drawn is limited to a small size, and visualization that satisfies the above requirements I couldn't. For example, in the conventional visualization method, there are visualization methods such as Cone Tree and Data Jewel Box, but when Cone Tree is used as the visualization method, the tree structure of the search tree can be drawn clearly, but each node is in the lower level. There is a problem that it is difficult to visualize the attributes included in the included data set so that the range of attribute values can be overlooked.

  It is possible to draw the attribute value range for the attributes included in the data set included in each node by changing the elements such as the color, thickness, and shape of the edges that indicate the relationship between the nodes. Only one attribute value range can be drawn per attribute.

  In addition, it is possible to draw the attribute value range for the attributes included in the data set included in each node on the lower figure by drawing the figure that represents the node. Since it is necessary to draw a graphic to be represented, and the size of the graphic representing the node is limited by that amount, it is difficult to draw the attribute value range for many attributes.

  Even if a data jewelry box is used as a visualization method, in addition to the tree structure of the search tree visualized two-dimensionally, it has a length in a direction orthogonal to the two-dimensional plane on which the graphic representing the data or node exists. It is only possible to draw the attribute value of one attribute included in the data or the attribute value aggregate value of one attribute included in the data set included in the lower level of each node in the rectangular parallelepiped or the cone. There was a problem that it was impossible to visualize the attribute value ranges of a plurality of attributes included in the included data set at the same time. The present invention solves these problems.

  In order to achieve the above object, a search tree drawing apparatus, a search tree drawing method, and a program according to the present invention represent a figure representing a node by a cone and a vertex of a cone representing a child node representing a parent node. By placing a cone that represents each node so that it touches the outer circumference of the bottom surface and does not touch other adjacent child nodes, the tree structure of the search tree is drawn in three dimensions, and information is drawn on the sides of the cone. To do.

Specifically, the retrieval tree drawing apparatus according to the present invention includes a data storage unit that holds a set of data including one or more attribute name / attribute value pairs in a tree structure, and each node of the tree structure is a cone. A cone drawing means for drawing by, a side information drawing means for drawing information on the side face of the cone, and a cone representing a child node among the nodes of the tree structure, a vertex of the cone representing the child node Is located on the outer periphery of the bottom surface of the cone representing the parent node among the nodes of the tree structure, and the outer periphery of the bottom surface of the cone representing the child node is disposed at a position where it does not contact with another adjacent child node possess a locating means, a search tree drawing means for performing recursively for each node and該錐body drawing means and side surface information drawing means and該錐body arranging means, wherein the side information drawing means, each node Data included in at least the search tree with respect to the side surface of the cone representing Related to the attribute associated with the region included in the side of the cone that represents each node, and the attribute of each attribute type is associated with each divided region in a one-to-one relationship. The attribute value associated with the region included in the data set included in the lower level is used as information relating to the attribute associated with the region rendered in the region included in the side of the cone that represents each node. Information obtained by normalizing the range between the maximum value and the minimum value of the attribute values associated with the region included in the entire data set including the range from the maximum value to the minimum value is drawn .

In the search tree drawing apparatus according to the present invention is to divide the number of regions of more attribute type number included in the data set a search tree including a side surface of the cone representing the nodes, each attribute to each area divided 1 By drawing information about attributes associated with each region in a one-to-one relationship, it is possible to overlook information about one or more attributes included in a data set included in each node at each node. Become.

In addition, the range of the maximum value to the minimum value of the attribute value associated with the area included in the data set included in each node in the lower side of each side of the cone representing the node is included in all the data sets included in the search tree. By drawing information normalized by the range of the maximum value and the minimum value of the attribute value associated with the region, the attribute value range for one or more attributes included in the data set included in each node is displayed as a node. It is possible to visualize so that it can be seen from above.

The retrieval tree drawing apparatus according to the present invention further includes a rotation angle calculation unit that calculates a rotation angle of the cone, and the side information drawing unit stores information drawn on a side surface of each cone. Redrawing may be performed in a state in which the rotation angle is rotated all at once around a perpendicular line perpendicular to each bottom surface from the vertex.
All of the information drawn on the side of each cone is redrawn at the same time by redrawing the information drawn on the side of each cone around the vertical line perpendicular to the bottom of each cone. Can be easily grasped and the information can be compared between nodes.

  In the retrieval tree drawing apparatus according to the present invention, the tree structure may be UBI-Tree.

Specifically, the retrieval tree drawing method according to the present invention includes a data storage means for holding a set of data including one or more attribute name / attribute value pairs in a tree structure, a cone drawing means, and side information drawing. A search tree drawing method executed by a search tree drawing device comprising: means and cone arrangement means, wherein the cone drawing means applies a cone to each node of the tree structure stored in the data storage means. A cone drawing procedure for drawing by a body, a side information drawing procedure for the side information drawing means to draw information on a side surface of the cone , and a cone arrangement means for child nodes of the nodes of the tree structure The vertex of the cone representing the child node is in contact with the outer periphery of the bottom surface of the cone representing the parent node among the nodes of the tree structure, and the outer periphery of the bottom surface of the cone representing the child node is Cone placement hand placed at a position not touching other adjacent child nodes If, have a, in the side information drawing procedure, dividing the side information drawing means, with respect to the side surface of the cone representing each node, the attribute number of types of areas included in the data set including at least the search tree Then, the attributes of each attribute type are associated with each divided area in a one-to-one relationship, information about the attributes associated with the area included in the side surface of the cone representing each node is drawn, and the cone representing each node As the information related to the attribute associated with the region to be drawn in the region included in the side of the node, the range from the maximum value to the minimum value of the attribute value associated with the region included in the data set included by the node is the search tree. The information normalized by the range of the maximum value and the minimum value of the attribute value associated with the region included in the entire data set included in is drawn .

In the search tree drawing method according to the present invention, the search tree drawing device further includes a rotation angle calculation unit, the rotation angle calculation unit calculates a rotation angle of the cone, and the side information drawing unit includes: A rotation processing procedure for redrawing the information drawn on the side surface of each cone in a state where the rotation angle is rotated all at once around a perpendicular line perpendicular to each bottom surface from the vertex of each cone,
Furthermore, you may have.

Specifically, in the search tree drawing program according to the present invention, the cone drawing means is stored in the data storage means for holding a set of data including one or more attribute name / attribute value pairs in a tree structure. The cone drawing procedure for drawing each node of the tree structure with a cone, the side information drawing means for drawing information on the side of the cone, and the cone arrangement means for drawing the cone structure A cone representing a child node of the node is connected to the outer periphery of the bottom surface of the cone representing the parent node of the tree structure node, and the cone representing the child node is connected. A search tree drawing program for causing a computer to execute a cone arrangement procedure for arranging a cone position at a position where the outer periphery of the bottom of the body does not contact other adjacent child nodes, and in the side information drawing procedure, the side information The drawing means is a cone representing each node A pyramid that divides the aspect into at least the number of attribute types included in the data set included in the search tree, associates the attribute of each attribute type in a one-to-one relationship with each divided area, and represents each node Draw information about the attribute associated with the region included in the side of the node, and the data included in the lower level as information regarding the attribute associated with the region rendered in the region included in the side of the cone representing each node The range from the maximum value to the minimum value of the attribute value associated with the region included in the set is normalized by the range of the maximum value and the minimum value of the attribute value associated with the region included in all data sets included in the search tree. A search tree drawing program characterized by drawing information .

  The above inventions can be combined as much as possible.

  According to the present invention, the shape of a node representing a node is represented by a cone, and the vertex of the cone representing a child node touches the outer periphery of the bottom surface of the cone representing the parent node and does not contact other adjacent child nodes. By arranging the cones representing the three-dimensionally, the tree structure of the search tree is drawn three-dimensionally, and information is drawn on the sides of the cones. This makes it possible to draw a large cone representing a node without using an edge, and to draw a lot of information on the side of the cone. Further, at this time, by using a cone as a figure representing a node and further performing the arrangement method, the relationship between the nodes can be expressed without using an edge.

The apparatus structure in this Embodiment is shown. The tree structure which manages the data set (120) which the data storage part (101) in this Embodiment hold | maintains is shown. The content of the data set (120) which the data storage part (101) in this Embodiment hold | maintains is shown. 7 shows a pointer table and an attribute value range table held by a tree structure for managing data held by the data storage unit (101) in the present embodiment. The flowchart of the search tree drawing process (step 300) which the search tree drawing part (114) in this Embodiment implements is shown. The cone drawing part (110) in this Embodiment shows the cone drawing image (400) which drawn the cone which represents a drawing object node with a cone by the cone drawing process (step 301). The side information drawing image (401) obtained by dividing the side surface of the cone representing the drawing target node into a plurality of regions by the side information drawing processing (step 302) by the side information drawing unit (111) in the present embodiment is shown. In the present embodiment, the side information drawing unit (111) further draws information on each side area of the cone representing nRT (200) as a drawing target node by the side information drawing process (step 302) (step 302). 402). In the present embodiment, the side information drawing unit (111) further draws information on each side area of the cone representing n1 (201) as a drawing target node by side information drawing processing (step 302) (step 302). 403). When the cone arrangement unit (112) in the present embodiment calculates the vertex of the cone representing the child node of the drawing target node by the cone arrangement processing (step 303), the calculated coordinates are used as the side information drawing image (404). It is shown above. In the present embodiment, a search tree drawing image (405) drawn by a search tree drawing process that recursively executes a drawing process is shown.

  Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiments described below are examples of the present invention, and the present invention is not limited to the following embodiments. In the present specification and drawings, the same reference numerals denote the same components.

(Device configuration of search tree drawing device)
FIG. 1 shows an apparatus configuration in the present embodiment.
The search tree drawing apparatus (100) includes a data storage unit (101). The data storage unit (101) holds the data set (120) using a tree structure. Further, the search tree drawing device (100) includes a search tree drawing unit (114), a cone drawing unit (110), a side information drawing unit (111), and a cone arrangement unit (112). Furthermore, the search tree drawing apparatus (100) includes a rotation angle calculation unit (113).

FIG. 2 shows a tree structure for managing the data set (120) held by the data storage unit (101) in the present embodiment.
In this embodiment, the tree structure for managing the data set (120) held by the data storage unit (101) is UBI-Tree. However, although the present embodiment is UBI-Tree, the present invention is not limited to this, and any search tree having a tree structure may be used. For example, the B-Tree or the R-Tree may be used.

  The tree structure that manages the data set (120) held by the data storage unit (101) is composed of three nodes nRT (200), n1 (201), and n2 (202). Holds data. Further, each node holds a pointer table and an attribute value range table shown in FIG. The data storage unit (101) holds a tree structure information table (210). The tree structure information table (210) has a maximum number of child nodes (211), an nRT height (212), and a pointer (213) to the nRT. ) And the rotation angle R.

  In the present embodiment, the search tree has a maximum number of child nodes (211) of 4, the height of the tree, that is, the height of nRT (212) is 2, and the rotation angle R is 90. It is not limited and may be larger or smaller. Also, the value of the pointer (213) to nRT is bfffdc18, but it can naturally change depending on the execution timing and execution environment.

FIG. 3 shows the contents of the data set (120) held by the data storage unit (101) in the present embodiment.
The data set (120) includes data (a) to (d). Data in the present invention is described in a data format composed of an arbitrary number of “key = value” sequences. At this time, “key = value” is synonymous with “attribute name = attribute value”. For example, data (A) is expressed as “x = 0, y = 5”, thereby expressing two attributes x and y and attribute values corresponding to the two attributes.

  Specific examples of the data referred to in the present invention include attributes such as temperature, humidity, current or voltage, fluid flow rate, substance concentration, brightness, noise, position, and acceleration. The attribute value is a value corresponding to each attribute. The present invention is not limited to this, and any data may be used as long as it has a value and can be expressed in the format of “attribute name = attribute value”. For example, it may be data acquired via the Web or the Internet. In the present embodiment, the number of data is four, but more data can be handled in the same manner.

FIG. 4 shows a pointer table and an attribute value range table held by a tree structure for managing data held by the data storage unit (101) in this embodiment.
In this embodiment, nRT (200) holds a pointer table (220) and an attribute value range table (230). At this time, the nRT pointer table (220) is a pointer to n1 (201) and n2 (202), and the attribute value range table (230) is an attribute value range table held by n1 (201) and n2 (202). Hold the attribute name and attribute value range. That is, the attribute value range table (230) held by nRT (200) is the same as the attribute value range table (231) held by n1 (201) "x = 0-5, y = 5-10", n2 (202 ) Held in the attribute value range table (232) is “x = 5 to 10, y = 0 to 5”, “x = 0 to 10, y = 0 to 10”.

  N1 (201) holds a pointer table (221) and an attribute value range table (231). At this time, the pointer table (221) of n1 (201) is a pointer to data (A) and data (B), the attribute value range table (231) is an attribute name held by data (A) and data (A), Holds the attribute value range. That is, in the attribute value range table (231) held by n1 (201), the data (a) held by n1 (201) is “x = 0, y = 5”, and the data (b) is “x = 5, When y = 10 ”,“ x = 0 to 5, y = 5 to 10 ”.

  Similarly, n2 (202) holds a pointer table (222) and an attribute value range table (232). At this time, the pointer table (222) of n2 (202) is a pointer to data (c) and data (d), and the attribute value range table (232) is an attribute name and attribute held by data (c) and data (d). Holds a range of values.

(Overall Flowchart in Search Tree Drawing Device)
FIG. 5 shows a flowchart of the search tree drawing process (step 300) performed by the search tree drawing unit (114) in the present embodiment. The retrieval tree drawing method according to the present embodiment includes an initialization process (step 501) and a drawing process (step 502) in this order.
First, as an initialization process (step 501), the nRT height (212) and the nRT pointer (213) are obtained from the tree structure information table (210) of the data storage unit (101), and the nRT height (212) is acquired. ) Is the height of the drawing target node, the pointer to nRT (213) is the pointer to the drawing target node, and the point (0, 0, 100) on the spatial coordinates (X, Y, Z) is the vertex coordinate of the drawing target node And Also, the number of rows in the attribute value range table (230) of nRT (200), that is, 2 is substituted for variable k, and a pointer to nRT is substituted for variable nRTp. Note that k is equal to the number of attribute types included in the tree structure. In this embodiment, the point (0, 0, 100) is the vertex coordinate of the drawing target node at the time of initialization processing. However, the present invention is not limited to this, and the vertex coordinate may be any coordinate.

Next, a rendering process is executed with the pointer to the initialized rendering target node, the vertex coordinates of the rendering target node, the height of the rendering target node, k, and nRTp as arguments (step 502).
In the drawing process,
First, the argument is passed to the cone drawing unit (110), and the cone representing the drawing target node is drawn by the cone drawing process (step 301).
Next, the argument is passed to the side information drawing unit (111), and information is drawn on the side of the cone representing the drawing target node by the side information drawing process (step 302).
Next, the argument is passed to the cone arrangement unit (112), and the vertex coordinates of the cone representing each child node of the drawing target node are calculated by cone arrangement processing.
However, if the height of the drawing target node is 1, that is, if the drawing target node is a leaf node, the next step 304 is skipped (step 303).

  Further, after reducing the height of the drawing target node by 1, drawing processing is executed for each child node of the drawing target node (step 304). At this time, the value reduced by 1 is the height of the drawing target node, the pointer to each child node is the pointer to the drawing target node, and the vertex of the cone representing each child node calculated in the cone arrangement processing (step 303) The coordinates are the vertex coordinates of the drawing target node, the variable k is the value as it is, that is, 2 and the variable nRTp is the value as it is, that is, the pointer to nRT, and these are the arguments. A pointer to each child node is obtained from the pointer table of the drawing target node.

Details of the cone drawing process (step 301), the side surface information drawing process (step 302), and the cone arrangement process (step 303) will be described later.
When drawing, using an existing drawing library such as DirectX or OpenGL, the information held by the data storage unit or the information calculated in the drawing process can be easily input into the drawing library. Can be created.
In the present embodiment, the drawing result may be displayed on a screen connected to the search tree drawing apparatus (100). Alternatively, the search tree drawing device (100) may be a Web application server, connected to a network, and a drawing image may be sent to a Web browser connected to the network to display a figure on the Web browser. .

  In the present embodiment, an example in which the drawing process (step 502) is performed in the order of steps 301 to 304 will be described, but the present invention is not limited to this. For example, if the base line of the drawing target node is determined in the cone drawing process (step 301), the side information drawing process (step 302) may be performed after the cone arrangement process (step 303). Further, the base line of the drawing target node is determined in the cone drawing processing (step 301), and the vertex coordinates of the cone representing each node calculated by the cone drawing processing (step 301) and the cone arrangement processing (step 303). If the height and the bottom radius are stored, the cone drawing process (step 301) and the cone arrangement process (step 303) are recursively executed for each node, and then the side information about each node is stored. Only the drawing process (step 302) may be recursively executed.

(Cone drawing unit 110)
FIG. 6 shows a cone drawing image (400) in which the cone drawing unit (110) in the present embodiment draws a cone representing a drawing target node by a cone drawing process (step 301). The cone drawing process (step 301) will be described with reference to FIG.

First, the cone drawing processing (step 301) is performed from the height of the drawing target node and the maximum number of child nodes (211) of the tree structure information table (210), from the height of the cone representing the drawing target node, that is, from the vertex coordinates. The length to the center point of the bottom of is calculated. In the present embodiment, the height of the cone is expressed by “the height of the cone representing the drawing target node of height b = 100 × ((1 / number of child nodes maximum value) ^ (nRT height−b))”. calculate. At this time, the height (212) of nRT is read from the tree structure information table (210).
Next, in the cone drawing process (step 301), the radius of the bottom surface of the cone representing the drawing target node is calculated. In the present embodiment, the radius of the bottom surface of the cone is equal to the height.

Next, in the cone drawing process (step 301), a cone representing a drawing target node is drawn according to the vertex coordinates of the cone, the height of the cone, and the radius of the bottom surface of the cone. As an example, drawing of a cone representing nRT (200) will be described.
First, in the cone drawing process (step 301), the height of a cone representing nRT (200) is calculated. Since the maximum value (211) of the child nodes is 4, the height (212) of the nRT is 2, and the height b of the drawing target node is also 2, the “cone representing the nRT (200) having the height 2” Height = 100 × ((1/4) ^ (2-2)) = 100 ”. At this time, since the vertex coordinates of the cone representing nRT (200) are (0, 0, 100), the center point of the bottom surface of the cone representing nRT (200) is (0, 0, 0).
Next, the cone drawing process (step 301) calculates the radius of the bottom surface of the cone representing nRT (200). In the present embodiment, the radius of the bottom surface of the cone is set to 100 because it is equal to the height.
Next, the cone drawing process (step 301) calculates a cone representing nRT (200) according to the vertex coordinates of the cone representing the given nRT (200), the calculated cone height, and the radius of the bottom of the cone. By drawing, the cone drawing image (400) shown in FIG. 6 is drawn.

  The present invention is not limited to this, and the height of each cone may be calculated by another method. For example, all the cones may have the same height, or the height of the cone representing the child node may be reduced to 1/10 of the height of the cone representing the parent node. The radius of the bottom surface of the cone is equal to the height, but the present invention is not limited to this and may be ½ of the height.

(Area division in the side information drawing unit 111)
FIG. 7 shows a side information drawing image (401) obtained by dividing the side surface of the cone representing the drawing target node into a plurality of regions by the side information drawing unit (111) in the present embodiment by the side information drawing processing (step 302). Show. The side information drawing process (step 302) will be described with reference to FIG.

  First, in the side information drawing process (step 302), k line segments are drawn on the side surface of the cone representing the drawing target node, the side surface of the cone is divided into k areas, and k types of attributes are assigned to each divided area. Are related in a one-to-one relationship. Specifically, the k line segments draw a straight line connecting the intersection of the center point of the bottom surface of the cone representing the drawing target node and the outer periphery of the bottom surface of the cone moved by the radius of the bottom surface of the cone in the X-axis direction. It is set as the base line of the target node, and is at an angle of “R degrees + (360 degrees / k) × n” (n = 0, 1, 2,..., K−2, k−1) from the base line on the bottom surface of the cone. A line segment connecting the intersection of the straight line and the circumference of the bottom surface and the apex of the cone. At this time, the rotation angle R is read from the tree structure information table (210).

  In the present embodiment, since k, that is, the number of rows in the attribute value range table (230) of nRT (200) is 2, two line segments are drawn on the side surface of the cone. Since the rotation angle R is 90 in the present embodiment, in nRT (200), only the center point (0, 0, 0) of the bottom surface of the cone representing nRT (200) and the radius of the bottom surface of the cone in the X-axis direction. The straight line connecting the intersection (100, 0, 0) with the outer periphery of the bottom surface of the moved cone is the base line (411) of nRT (200), and the first line segment, that is, the line segment of n = 0 is the base line (411). Is a line segment connecting the point (0, -100, 0) located at "90 degrees" to the apex (0, 0, 100) of the cone, and the second line segment is "90 degrees + (360 degrees / 2) × 1 ”, that is, a line segment connecting the point (0, 100, 0) located at“ 270 degrees ”from the base line (411) and the apex (0, 0, 100) of the cone. Thus, the side information drawing image (401) shown in FIG. 7 is drawn.

  In the present embodiment, the areas are divided so that the sizes thereof are equal. However, the present invention is not limited to this, and the areas may be divided into different sizes. Further, it may be divided into k or more regions.

  Next, the side information drawing process (step 302) associates the areas divided by the k line segments with k attributes. In the present embodiment, among the regions divided by the two line segments, the first line segment, that is, the region from the n = 0 line segment to the second line segment, that is, the n = 1 line segment is defined as the attribute x. Similarly, the area from the second line segment to the first line segment is associated with the attribute y. The present invention is not limited to this, and each region may be associated with each attribute in reverse order.

(Information drawing in the side information drawing unit 111)
8 and 9, the side information drawing unit (111) in the present embodiment further draws information in each area on the side of the cone representing the drawing target node by the side information drawing processing (step 302). The side information drawing images (402) and (403) are shown. The side information drawing process (step 302) will be described with reference to FIG.

  The side information drawing process (step 302) refers to the attribute value range table (230) of nRT (200) and the attribute value range table of the drawing target drawing target node, and for each attribute, the attribute value range table (nRT (200)) 230), that is, the attribute value range R1 in all data sets held by the tree structure and the attribute value range described in the attribute value range table of the drawing target node, that is, held below the drawing target node The attribute value range R2 in the data set to be acquired is acquired, and the range of R2 with respect to R1 of the associated attribute is drawn in the area. In this embodiment, the point in contact with the bottom of the cone in the region is the starting point of the range of R1, the point that is the apex of the cone is the end of the range of R1, and the range occupied by R2 in FIG. Draw with diagonal or dotted shading as shown.

As an example, a drawing of a cone representing nRT (200) and n1 (201) will be described.
In the cone representing nRT (200) in this embodiment, for attribute x, R1 is x = 0 to 10 and R2 is also x = 0 to 10, so the region associated with attribute x The whole range is drawn with hatched shading. As for the attribute y, since R1 is y = 0 to 10 and R2 is similarly y = 0 to 10, the entire range in the region associated with the attribute y is drawn with a dotted line. Thereby, the side information drawing image (402) shown in FIG. 8 is drawn.

  When n1 (201) is a drawing target, for attribute x, R1 is x = 0 to 10, R2 is x = 0 to 5, that is, R2 is 0% to 50% with respect to the range of R1. Since this is the range of the part, in the region associated with the attribute x, a range of 0 to 50% from the side closer to the bottom surface is drawn by hatching. For attribute y, R1 is y = 0 to 10 and R2 is y = 5 to 10, that is, R2 is in the range of 50% to 100% with respect to the range of R1, and therefore is associated with attribute y. In this region, a range of 50% to 100% from the side closer to the bottom surface is drawn by dotted line shading. As a result, the side information drawing image (403) shown in FIG. 9 is drawn.

When R1 is 0 to 10 and R2 is 2 to 4, a range of 20% to 40% from the side closer to the bottom surface is drawn with hatched or dotted lines.
However, the present invention is not limited to this, and the point that touches the apex of the cone may be the starting point of the range of R1, and the point that becomes the bottom of the cone may be the end of the range of R1, and the shading used for drawing the range is related to the attribute. They may be unified with diagonal lines or dotted lines. Also, different colors may be used for drawing the range for each attribute, and the colors may be made translucent. Further, the drawing may be performed by changing the range of R2 with respect to R1 depending on the color intensity or luminance.

  In this embodiment, normalization according to the present invention is included in all data sets included in the search tree, as described above, in this embodiment, the attribute value range of the attribute associated with the region in which the drawing target node is included in the data set below is included. This refers to conversion to a range occupied in the attribute value range of the attribute associated with the region.

(Cone arrangement part 112)
FIG. 10 shows the calculated coordinates when the cone arrangement unit (112) in the present embodiment calculates the vertex of the cone representing the child node of the drawing target node by the cone arrangement processing (step 303). This is shown on the image (404). The cone arrangement process (step 303) will be described with reference to FIG.

  First, the cone arrangement unit (112) uses a base point of the base of the cone representing the drawing target node and the outer intersection as a base point, and uses this base point as a vertex coordinate of the cone representing the first child node. For example, when the drawing target node is nRT (200), the base line (411) and the intersection (100, 0, 0) of the outer circumference are the base points.

  Next, the cone arrangement part (112) divides the central angle 360 degrees of the bottom surface of the cone by the number of child nodes, and the straight line extending from the center point of the bottom surface of the cone toward the outer periphery at the divided angle and the bottom surface of the cone Let the intersection of the outer peripheries be the vertex coordinates of each child node. At this time, the number of child nodes is calculated from the number of rows in the pointer table of the drawing target node. In the cone representing nRT (200) in this embodiment, nRT (200) has two child nodes, n1 (201) and n2 (202), and the pointer table (220) has two rows. The center angle 360 degrees of the bottom surface of the cone is divided by two. In this case, n1 is a base point (100, 0, 0) on the base line (411), and n2 is an intersection with the outer periphery when the base line (411) is rotated 360 degrees / 2 from the base point (100, 0, 0) ( −100, 0, 0) is the vertex coordinate.

  By performing the above drawing processing recursively, the entire search tree can be drawn. The search tree drawn in this way expresses the tree structure of the search tree in a three-dimensional manner by using a cone representing each node and its arrangement without using an edge representing the relationship between the nodes. Information related to the attribute related to the area is expressed on each area of the side surface of the cone to be represented, and further, the information related to the attribute related to the area is included in the data set included in the lower level of the node corresponding to the cone The attribute value range of the attribute is expressed, and this has a feature that it is possible to overlook which attribute value range is included in the data set included in each node at the lower level.

(Completed drawing)
FIG. 11 shows a search tree drawing image (405) drawn by the search tree drawing process in which the drawing process is recursively performed in the present embodiment. The search tree drawn in the present embodiment does not use an edge representing the relationship between the nodes as described above, and uses the cones representing the nodes nRT (200), n1 (201), and n2 (202) and the arrangement thereof. Represents the tree structure of the search tree three-dimensionally, and further expresses information related to the attribute x and attribute y related to the region on each side surface of the cone representing each node, and further relates to the region Represents the attribute value range of the attribute x and the attribute y included in the data set included in the lower level of the node corresponding to the cone as information on the attribute x and the attribute y, and thereby the data set included in the lower level of each node It is possible to overlook how much attribute value range the attribute x and the attribute y are included in.

  In this embodiment, the search tree drawing unit (114) recursively draws from the root node. However, the present invention is not limited to this, and the search tree drawing unit (114) recursively starts from any inner node. Only the subtree below the inner node may be drawn. In addition, if a pointer from a child node to a parent node is also managed by each child node, other nodes can be drawn starting from a leaf node.

(Rotation processing using the rotation angle calculation unit 113)
Further, the retrieval tree drawing apparatus (100) uses the drawing image shown in FIG. 11 to center the information drawn in each area on the side surface of the cone representing each node from the vertex of each cone to the bottom surface. The rotation processing may be executed to redraw the side information drawing image obtained by rotating each cone by the angle calculated by the rotation angle calculation unit (113).

The rotation processing procedure in this embodiment is shown below.
After the search tree drawing unit (114) performs the search tree drawing process (step 300), the rotation angle calculation unit (113) performs the following process.
The rotation angle calculation unit (113) updates the value of the rotation angle R in the tree structure information table (210) to a value increased by 90 from Δ at regular intervals, and further searches the search tree drawing unit (114). A drawing process (step 300) is called.

  In the present embodiment, Δ is 5, and R after a certain period of time is “R = 90 + 5”, that is, 95. Furthermore, the rotation angle R of the tree structure information table (210) after a certain period of time is “R = 95 + 5”, that is, 100. In this embodiment, 3 seconds is set as the predetermined period. Then, in this embodiment, the information to be drawn in each area is rotated by 5 degrees and redrawn every 3 seconds.

  However, the present invention is not limited to this, and the image may be redrawn by being rotated every second, or may be redrawn by receiving an operation command from the user and rotating at the timing when the operation command is received. Further, the rotation angle Δ may take a negative value, and the information drawn in each area may be reversely rotated and redrawn.

  The rotation process is calculated by the cone drawing process (step 301) and the cone arrangement process (step 303) for the tree structure drawn by the search tree drawing unit (114) performing the search tree drawing process (step 300). By storing the vertex coordinates, height, and bottom surface radius of the cone representing each node, only the side information rendering process (step 302) is recursively performed for each node, and information is rendered in each region. May be redrawn.

  Thus, by rotating and redrawing information drawn in each area on the side surface of the cone representing each node, any area can be drawn on the front surface. As a result, it is possible to easily grasp all information drawn on the side surface of each cone, and it is possible to compare all information between nodes.

  The present invention can be applied to the information communication industry.

100: Search tree drawing device 101: Data storage unit 110: Cone drawing unit 111: Side surface information drawing unit 112: Cone arrangement unit 113: Rotation angle calculation unit 114: Search tree drawing unit 120: Data set 200: nRT
201: n1
202: n2

Claims (6)

Data storage means for holding a set of data including one or more pairs of attribute names and attribute values in a tree structure;
A cone drawing means for drawing each node of the tree structure by a cone;
Side information drawing means for drawing information on the side of the cone;
A cone representing a child node of the nodes of the tree structure, and a vertex of the cone representing the child node is in contact with an outer periphery of a bottom surface of the cone representing a parent node of the nodes of the tree structure; A cone arrangement means arranged at a position where the outer periphery of the bottom surface of the cone representing the node does not contact with other adjacent child nodes;
Search tree drawing means for recursively executing the cone drawing means, the side surface information drawing means, and the cone arrangement means for each node;
I have a,
The side information drawing means includes:
The sides of the cone representing each node are divided into at least the number of attribute types included in the data set included in the search tree, and the attributes of each attribute type are associated with each divided region in a one-to-one relationship. , Draw information about the attributes associated with the region that the side of the cone representing each node contains,
As information about the attribute associated with the area drawn in the area included in the side surface of the cone representing each node, the maximum value to the minimum value of the attribute value associated with the area included in the data set included in the node below An apparatus for drawing a search tree, wherein the information is normalized by the range of the maximum value and the minimum value of attribute values associated with the area included in all data sets included in the search tree. A rotation angle calculating means for calculating a rotation angle of the cone;
The side surface information drawing means redraws the information drawn on the side surface of each cone in a state in which the rotation angle is simultaneously rotated around a perpendicular line perpendicular to each bottom surface from the vertex of each cone. The retrieval tree drawing apparatus according to claim 1 , wherein: Search Tree drawing apparatus according to claim 1 or 2 the tree structure is characterized by a UBI-Tree. Retrieval tree drawing apparatus comprising: data storage means for holding a set of data including one or more attribute name / attribute value pairs in a tree structure ; cone drawing means; side information drawing means; and cone arrangement means. Is a search tree drawing method executed by
A cone drawing procedure in which the cone drawing means draws each node of the tree structure stored in the data storage means by a cone;
A side information drawing procedure in which the side information drawing means draws information on a side surface of the cone;
The pyramid arrangement means is configured to display a cone representing a child node of the nodes of the tree structure, and a vertex of the cone representing the child node of a bottom surface of the cone representing a parent node of the nodes of the tree structure. A pyramid placement procedure that touches the outer circumference and places the outer circumference of the bottom surface of the cone that represents the child node at a position that does not touch other adjacent child nodes;
I have a,
In the side information drawing procedure, the side information drawing means includes:
The sides of the cone representing each node are divided into at least the number of attribute types included in the data set included in the search tree, and the attributes of each attribute type are associated with each divided region in a one-to-one relationship. , Draw information about the attributes associated with the region that the side of the cone representing each node contains,
As information about the attribute associated with the area drawn in the area included in the side surface of the cone representing each node, the maximum value to the minimum value of the attribute value associated with the area included in the data set included in the node below The information normalized by the range of the maximum value and the minimum value of the attribute value associated with the region included in the entire data set included in the search tree is drawn.
A search tree drawing method characterized by that . The search tree drawing device further includes a rotation angle calculation means,
The rotation angle calculation means calculates the rotation angle of the cone, and the side information drawing means perpendicularly draws the information drawn on the side surface of each cone from the apex of each cone to each bottom surface. A rotation processing procedure for redrawing in a state where the rotation angle is rotated all at once around the center,
The retrieval tree drawing method according to claim 4 , further comprising: The cone drawing means draws each node of the tree structure stored in the data storage means that holds a set of data including one or more attribute name / attribute value pairs in a tree structure by the cone. Procedure and
A side information drawing means for drawing information on a side surface of the cone;
The cone arrangement means is a cone representing a child node among the nodes of the tree structure, and an outer periphery of a bottom surface of the cone whose apex of the cone representing the child node represents a parent node of the nodes of the tree structure A cone placement procedure that touches the top and places the outer periphery of the bottom surface of the cone representing the child node at a position that does not touch other adjacent child nodes;
A search tree drawing program for causing a computer to execute the,
In the side information drawing procedure, the side information drawing means includes:
The sides of the cone representing each node are divided into at least the number of attribute types included in the data set included in the search tree, and the attributes of each attribute type are associated with each divided region in a one-to-one relationship. , Draw information about the attributes associated with the region that the side of the cone representing each node contains,
As information about the attribute associated with the area drawn in the area included in the side surface of the cone representing each node, the maximum value to the minimum value of the attribute value associated with the area included in the data set included in the node below The information normalized by the range of the maximum value and the minimum value of the attribute value associated with the region included in the entire data set included in the search tree is drawn.
A search tree drawing program characterized by that .

Images