AU2004237905A1 - Method of generating a style family from template examples - Google Patents

Description

S&F Ref: 687347

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT Name and Address of Applicant: Actual Inventor(s): Address for Service: Canon Kabushiki Kaisha, of 30-2, Shimomaruko 3-chome, Ohta-ku, Tokyo, 146, Japan Lena Qian Spruson Ferguson St Martins Tower Level 31 Market Street Sydney NSW 2000 (CCN 3710000177) Method of generating a style family from template examples Invention Title: The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5845c -1- METHOD OF GENERATING A STYLE FAMILY FROM TEMPLATE EXAMPLES Copyright Notice This patent specification contains material that is subject to copyright protection.

O The copyright owner has no objection to the reproduction of this patent specification or C1 related materials from associated patent office files for the purposes of review, but otherwise reserves all copyright whatsoever.

Field of the Invention The present invention relates to the presentation of data in a computer environment.

In particular, the present invention relates to the method of generating a style family from a group of presentation templates that may be in different formats.

Background In a computer environment, data can be presented in many different formats, e.g., bar chart, table, tree or just text. A graphical presentation template is used to set up the format and styles of the data to be presented. A graphical presentation template, hereinafter simply a 'template', typically defines graphical objects. Some of the graphical objects are directly related to the data to be presented, and others may be indirectly related or just decorative. For example, bars in a bar chart will typically be rectangular objects having a specified fill, often an opaque colour, but sometimes having some transparency, representing a pattern (eg. checked or striped, or the like). Other objects such as lines may be used to represent axes of the chart, and such may be defined by colour and stroke width for example. It is the combinations of objects and their respective styles that afford an overall appearance to a template which may be recognisable and hopefully pleasing to the eye of a viewer. Data presentations produced from the same template have the same style, and thus the same visual effects, and can 687347 -2results in carrying a common stylistic theme amongst a number of presentations (eg. a presentation of financial results including tables, bar charts, pie charts and graphs.

A template 'family' groups together one or more templates that may be of t different formats, but which are of the same or have a consistent style. For example, a family of templates for bar chart, pie chart, plot and table can be defined using similar NI shapes, colours from the same colour scheme and the same font style. As the styles for these different templates share some commonalities, they can be used together to achieve a consistent visual effect.

Usually, a template is associated with a set of styles. In the Swiff Chart Pro T M software application, manufactured by GlobFX Technologies SARL of France, the user can pick a style for a selected chart type, such as bar chart or pie chart, from a range of named styles grouped as "Paris", "San Francisco", "Classic-Sand", etc. Those different charts using the same named style have the same look and feel. However, when a user changes to, for example, a "Classic-Sand" style on an existing bar chart, the change cannot be carried over to another pie chart in the same group. In other words, the styles are not shared within a group of charts. When a new family of templates is required, the graphic designer needs to design this new set of templates individually. The graphic designer may try to keep the styles consistent across the templates in the family by copying and pasting the colours from one template to another template. This process is tedious and inefficient.

To be able to share styles across different templates, some software applications use style sheets. For example, Cascading Style Sheets (CSS) can be applied on a HTML document. Microsoft Word T M 2002, manufactured by Microsoft Corporation of the USA, can define a heading style for outline levels in a document. Similarly, Windows XP T M operating system, also manufactured by Microsoft Corporation, has an appearance setting 687347 -3o for any active/inactive windows and message boxes shown in the graphical user interface thereof. The sharing of styles can therefore be used for templates with similar formats such as different kinds of bar charts, but it cannot be easily applied to templates of t different formats such as a table, a bar chart and a plot.

S 5 To overcome this problem, the Extensible Stylesheet Language (XSL) program e¢3 can be used. XSL consists of two parts: a method for transforming Extensible Markup Language (XML) documents, and a method for formatting XML documents using formatting semantics for styles, layout, etc. Generalised styles can be defined in the style sheet. An XSL program can transform references within graphical objects to the styles in the style sheet. However, using XSL requires the graphic designer to anticipate all the different formats of the templates. If a template is a bar chart template, the XSL program knows which graphical objects will use a certain style defined in the style sheet. In this case, in order to have new styles the graphic designer does not need to design a set of new templates from scratch as long as the new styles can be derived from existing templates via XSL programs. However, programming in XSL requires extensive training and experience, especially when a new template is to be added to a template family.

Another way of creating styles for a family of templates involves manually extracting styles from existing templates and adjusting them to fit each individual template as well as across the templates. This process is iterative and time consuming.

When more than one graphic designer is involved in designing a set of templates, it is more difficult to maintain the consistency of the styles across the templates. Thus, a method is required to allow easy and efficient integration of a new template into a style family.

US Patent Application Publication No. US2004/0034613A1 describes an arrangement that can generate style properties from a document layout as a constraint 687347 -4satisfaction problem comprising a set of variables, values and constraints. The variables are font parameters or text line parameters. When the same style parameters are represented for different output devices, the values of the variables are changed following the same constraints. Although the styles are generated and represented in a different 0 way, the document layout using the styles is the same. In other words, the styles cannot be applied to different documents, but only to the same document targeted at different Senvironments, mobile cellular telephone or personal digital assistant (PDA) device.

Thus, this method cannot be applied to the generation of a style family.

Generating style sheets using data mining from sample HTML documents and CSS code is one of the techniques used to classify styles and types of web-based information.

The style sheet in CSS is generated for one set of styles from similar HTML documents.

If a different HTML document has a similar style, it refers to the same CSS style sheet.

This method is useful to identify similar styles from different templates, but it does not serve the purpose of generating the styles for a family of templates.

Summary It is an object of the present invention to substantially overcome, or at least ameliorate, one or more disadvantages of existing arrangements for creating templates to be used to graphically present data.

In accordance with a first aspect of the present disclosure there is provided a method of generating a style family and virtual template from existing templates, the existing templates being arranged in different formats, said method comprising the steps of: 687347 extracting styles and shapes from at least one said existing template into style and shape statistic tables by analysing each graphical object in the said one existing template; obtaining one or more dominant styles from said style and shape statistic tables, said dominant style being a style shared by one or more graphical objects that cover prominent areas of a space of said one existing template; adapting said dominant styles from the style and shape statistic tables to thereby generate said style family; and forming a virtual template based on said one existing template, said virtual template having at least one graphical object referring to styles in said style family.

In accordance with a second aspect of the present disclosure there is provided a method of generating a template from a style family and a virtual template formed according to the first aspect of described above, said method comprising the steps of: obtaining dominant style names from said virtual template; comparing said dominant style names with the dominant styles in the style family; and if the style names match the dominant styles, modifying the graphical objects in said virtual template referring to the dominant styles with a style specified in the style family.

According to a third aspect of the present disclosure there is provided a method of forming a representation of present data, said method comprising: analysing shapes of graphical objects present in at least one existing representation of data to form at least one template for the representation of data; 687347 C -6analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; selecting at least one said template and applying one said style family to graphical objects thereof to form a stylized template; and i applying said present data to said stylized template to thereby form a Srepresentation of said present data.

According to a fourth aspect of the present disclosure there is provided a method of forming a representation of present data, said method comprising the steps of: analysing shapes of graphical objects present in at least one existing representation of data to form at least one template for the representation of data; selecting, from a library of style families of graphical objects, a style family desired to be used in representing said present data; selecting at least one said template and applying said selected style family to graphical objects thereof to form a stylized template; and applying said present data to said stylized template to thereby form a representation of said present data.

According to a fifth aspect of the present disclosure there is provided a method of forming a representation of present data, said method comprising: analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; selecting, from a library of templates of graphical objects structured for the representation of data, a template for the representation of said present data; 687347 -7selecting at least one said style family and applying said selected style family to graphical objects of said selected template to form a stylized template; and applying said present data to said stylized template to thereby form a t representation of said present data.

Other aspects, including computer programs and apparatus are also disclosed.

C Brief Description of the Drawings One or more embodiments of the present invention will now be described with reference to the drawings, in which: Fig. 1 is a schematic block diagram of a general purpose computer upon which arrangements described can be practiced; Fig. 2 illustrates a workflow of creating templates and generating style families; Fig. 3 is an example of a bar chart template; Fig. 4 shows partial data of an example for a bar chart template in SVG; Fig. 5A illustrates an example of input and output of the style family generator; Fig. 5B is an example of input and output of the template generator; Fig. 6 depicts the main process of generating a style family and a virtual template from a template; Fig. 7 is a flowchart of extracting styles; Fig. 8 shows an outline of a virtual template for the bar chart template; Fig. 9 is a flowchart to obtain dominant styles; Fig. 10 schematically illustrates the generation of a new representation from shape and style analysis of existing representations; Fig. 11 is an example of a dominant style list; Fig. 12 shows an example of a style family; Fig. 13A is a flowchart to create a style family from the first template in a 687347 -8- Stemplate family; d Fig. 13B is a flowchart to adapt styles of a template to a style family; Fig. 14 depicts a flowchart to generate a virtual template from a template; t Fig. 15 is an example of a virtual template in SVG; 5 Figs. 16A and 16B show variations on the approach of Fig. 10 according firstly to Cca library of dominant styles and secondly for a library of shape templates; Fig. 17 is a flowchart of generating a template from a style family and a virtual template; Detailed Description Including Best Mode The methods of creating templates to be described are preferably practiced using a general-purpose computer system 100, such as that shown in Fig. 1 wherein the processes of Figs. 2 to 17 may be implemented as software, such as an application program executing within the computer system 100. In particular, the steps of template creation are effected by instructions in the software that are carried out by the computer. The instructions may be formed as one or more code modules, each for performing one or more particular tasks. The software may also be divided into two separate parts, in which a first part performs the template creation methods and a second part manages a user interface between the first part and the user. The user interface is typically formed by a graphical user interface (GUI). The software may be stored in a computer readable medium, including the storage devices described below, for example. The software is loaded into the computer from the computer readable medium, and then executed by the computer. A computer readable medium having such software or computer program recorded on it is a computer program product. The use of the computer program product 687347 -9in the computer preferably effects an advantageous apparatus for template creation according to the present disclosure.

The computer system 100 is formed by a computer module 101, input devices t such as a keyboard 102 and mouse 103, output devices including a printer 115, a display 0 5 device 114 and loudspeakers 117. The display device 114 is used to implement the GUI, Stypically in concert with the keyboard 102 and mouse 103. A Modulator-Demodulator (Modem) transceiver device 116 is used by the computer module 101 for communicating to and from a communications network 120, for example connectable via a telephone line 121 or other functional medium. The modem 116 can be used to obtain access to the Internet, and other network systems, such as a Local Area Network (LAN) or a Wide Area Network (WAN), and may be incorporated into the computer module 101 in some implementations.

The computer module 101 typically includes at least one processor unit 105, and a memory unit 106, for example formed from semiconductor random access memory (RAM) and read only memory (ROM). The module 101 also includes an number of input/output interfaces including an audio-video interface 107 that couples to the video display 114 and loudspeakers 117, an 1/O interface 113 for the keyboard 102 and mouse 103 and optionally a joystick (not illustrated), and an interface 108 for the modem 116 and printer 115. In some implementations, the modem 11l6 may be incorporated within the computer module 101, for example within the interface 108. A storage device 109 is provided and typically includes a hard disk drive 110 and a floppy disk drive 111. A magnetic tape drive (not illustrated) may also be used. A CD-ROM drive 112 is typically provided as a non-volatile source of data. The components 105 to 113 of the computer module 101, typically communicate via an interconnected bus 104 and in a manner which results in a conventional mode of operation of the computer 687347 o system 100 known to those in the relevant art. Examples of computers on which the described arrangements can be practised include IBM-PC's and compatibles, Sun Sparcstations or alike computer systems evolved therefrom.

t Typically, the application program is resident on the hard disk drive 110 and read 5 and controlled in its execution by the processor 105. Intermediate storage of the program NI and any data fetched from the network 120 may be accomplished using the semiconductor memory 106, possibly in concert with the hard disk drive 110. In some instances, the application program may be supplied to the user encoded on a CD-ROM or floppy disk and read via the corresponding drive 112 or 111, or alternatively may be read by the user from the network 120 via the modem device 116. Still further, the software can also be loaded into the computer system 100 from other computer readable media. The term "computer readable medium" as used herein refers to any storage or transmission medium that participates in providing instructions and/or data to the computer system 100 for execution and/or processing. Examples of storage media include floppy disks, magnetic tape, CD-ROM, a hard disk drive, a ROM or integrated circuit, a magneto-optical disk, or a computer readable card such as a PCMCIA card and the like, whether or not such devices are internal or external of the computer module 101. Examples of transmission media include radio or infra-red transmission channels as well as a network connection to another computer or networked device, and the Internet or Intranets including e-mail transmissions and information recorded on Websites and the like.

Fig. 2 is a workflow of creating templates in a data presentation system which may be implemented within the computer system 100 and represented to a user, such as a graphic designer, via the GUI on the display 114. Usually the graphic designer 210 will design and modify templates using an authoring tools process 220. A template 250 is formed of two groups of graphical objects: one for the presented data, referred to as data 687347 -11- Sgraphical objects bars in a bar chart) and the rest for template decorations, called non-data graphical objects XY axes in a bar chart). While the data graphical objects present sample data (ie. a sample length of the bars in bar chart), non-data graphical Sobjects are decorative and maintain the same look with the possibility of minor changes 0 5 depending indirectly on the data.

NI A graphical object can be associated with a set of styles, colour and font. If the graphical object is a shape, its style can also include further properties of the shape.

For example, a rectangle with round comers may specify a radius parameter for the curvature of the corners. Fig. 3 illustrates an example of a bar chart template. The data graphical objects are labelled from 302 to 306 and from 310 to 328. The rest of the graphical objects are non-data graphical objects including XY axes 340 and 360, a percentage sign 370, scales 0 to 100 on the Y axis, a caption 365 and a background 380.

In a preferred implementation, the template is created using SVG (W3C Scalable Vector Graphics 1.1 Recommendation). This recommendation defines a vector graphics language and can be used to represent a data presentation involving vector graphical constructs such as lines, curves, rectangles, etc. The Recommendation is expressed using the W3C Recommendation for Extensible Markup Language (XML).

Fig. 4 shows a partial SVG description of the bar chart template in Fig 3. The SVG description demonstrates a hierarchical structure in which data graphical objects are grouped separately from the non-data graphical objects.

During the template design authoring process 220, the graphic designer creates a set of graphical objects and adds their styles. All templates in a family 280 share some common styles. Sometimes, the designer copies a colour or a font from one template to another template. The styles are associated with the graphical objects in the templates 250. When a template 250 is removed from the family, the styles are lost.

687347 -12- To overcome this problem, the graphic designer 210 can use a method described herein to generate style families and generate templates. Preferably, the method is implemented in a software application 200 that includes two modules: a style family Sgenerator 230 and a template generator 240. The style family generator 230 takes templates 250 and produces style families 260 and virtual templates 270. The virtual (Ni N, templates 270, are similar to the input template 250 except referencing to the styles defined in the style family 260. The graphic designer can also use the template generator 240 to generate templates 250 from selected style families 260 and virtual templates 270. Any generated template can then be modified by the graphic designer using the authoring tools 220. Once the template is modified, it can be input to the style family generator 230 to consolidate the style family.

Fig. 5A shows an example of the input and output of the style family generator 230. A style family contains the styles that can be shared by the templates in a template family. Three templates, one for each of a bar chart 520, pie chart 522 and table 524, belong to a family and another plot template 526 is designed for another family The templates 520, 522, 524 and 526 are designed by the graphic designer 210 in an authoring tool (eg. 220). The bar chart template 520 is firstly used to generate the style family 540. Then the pie chart 522 and table templates and 524, are used to integrate their styles into the style family 540. Three virtual templates 550 to 554 are then created based on the original templates 520 to 524. These virtual templates reference to the style family 540. Another template family that contains only one plot template 526 can be input to the style family generator to produce another style family 546 and a corresponding virtual template 556.

Fig. 5B illustrates an example of generating templates using the template generator 240. To avoid designing the templates for pie chart, bar chart and table in 687347 -13another template family from scratch using the authoring tool 220, the graphic designer can select a style family 546 and a set of virtual templates 550 to 554 to generate new templates 560 to 564. The generated templates for the style family can be then modified by the authoring tool 220 as necessary.

Fig. 6 is a data and process flow diagram illustrating the actions 600 taken by the style family generator 230. An initial process of extracting styles 620 takes a template610 and extracts information from graphical objects contained in the template 610. The extracted information is stored into style and shape statistic tables 630.

A process 640 then analyses the style and shape statistic tables so as to form dominant style lists 650.

The dominant style lists 650 comprise the styles used by the graphical objects that cover visually significant areas of the template space. As such, those styles tend to dominate the perception of a viewer. Further, whilst a first style may well be more prominent than a second style, the second style may well be substantially more dominant than a third or other styles. Such a template may therefore have two dominant styles.

Where a number of styles are substantially equally prominent, all such styles may be classed as "dominant". Specific quantitative measures for assessing dominance are discussed later.

If the template is the first one to be added to a template family, a process 660 then creates a style family 680 by copying the dominant styles from the lists 650 of the template. If the style family 680 is already generated and the template 610 is another member of the template family, the process 660 integrates the dominant style lists 650 into the style family 680. Finally, a process 690 creates a virtual template 695 that has references to the style family 680.

Fig. 7 is a flowchart illustrating the "extract styles" process 620. The process 620 687347 -14executes the steps in Fig. 7 a first time to extract styles for the data graphical objects and then a second time to process the non-data graphical objects. Each execution of Fig. 7 outputs a style statistic table and a shape statistic table.

The process 620 starts at step 720, in which a style statistic table 630a and a shape statistic table 630b are initialised. Each row in the style statistic table 630a corresponds Cc- I to one style property, colour or font. While a colour defines an attribute of an object, a font style combines a set of properties such as weight, family, italic variation and smallcap variation. Each row in the shape statistic table 630b is defined for each shape graphical object type with its layout features. The position and size of the shape are not included in the row. For example, an entry in the shape statistic table may be a rectangle with zero radius of the rounded corner.

Each row in the style statistic table 630a contains a style and statistics that describe how the style is used in the template 610. Preferably, the statistics include the following: 1. Number of graphical objects sharing the style; 2. A list of graphical objects sharing the style; 3. Coverage percentage in the total space covered by the template; and 4. A list of colours for the neighbouring or adjacent objects if the style is color.

Similarly, each row in the shape statistic table 630b contains a shape type and statistics that describe how the shape is used in the template 610. Preferably, the statistics include the following: 1. Number of graphical objects using the same shape type and layout features; 2. A list of graphical objects sharing the shape type and layout features; and 3. Coverage percentage in the total space covered by the template.

At step 730, a graphical object is obtained and its styles are extracted from the 687347 o template 610. Step 740 continues to get a style from the styles extracted in step 730 and d compares the style with the existing styles in the style statistic table 630a. Individual properties (eg. colour, font type, font size) of the particular styles are compared. If no match between any of the properties is found at decision step 750, a new row with the 5 style is added to the style statistic table 630a at step 770. Then the graphical object is C added to the created row of the table and its corresponding statistics are calculated and filled in the row of the table at step 760. If a match is found in step 750 between the style and a style in the style statistic table 630a step 760 is executed. The graphical object is then added to the row with which the style is matched and the corresponding style statistics in the row of the table 630a are updated.

A check is made at step 765 to determine whether the graphical object contains more styles to be analysed. If so, the execution returns to step 740. Otherwise, step 783 looks for a match for the graphical object in the shape statistic table 630b. If no matching shape is found in this table, a new row for the shape is added to the shape statistic table 630b at step 786. Then the process executes step 788 to update the statistics in the row. If a shape is found in the shape statistic table 630b, step 788 is executed and the matching shape's statistics are updated.

At step 785, a check is made to determine whether any more graphical object is to be processed. If so, the process returns to step 730. Otherwise, the process terminates at step 799.

An example of the style statistic table 630a for the non-data graphical objects in the bar chart template in Fig. 3 is shown in Table 1 below.

Table 1: Style statistic table for the non-data graphical objects 687347 -16- Style Id Styles Num. Graphical Coverage Adjacent of objects to overall colour list objects Style -1 Colour 000000 1 Text for 1 D6D6D7, caption 3A3A38 Style -2 Font: size=14, 1 Text for weight =bold Caption Style -3 Colour =D6D6D7 1 background Style -4 Colour =3A3A38 2 X and Y axes 20 D6D6D7, 8C102D Style -5 Colour =8C102D 1 X axis line 2 D6D6D7, 3A3A38 Style -6 Colour =FFFFFF 11 Text for Y axis 10 3A3A38 scales Style-7 Font: size=12, 11 Text for Y axis weight=plain scales The shape statistic table 630b for the non-data graphical objects is shown in Table 2. The only shape defined in the bar chart of Fig. 3 is rectangle.

Table 2: Shape statistic table for the non-data graphical objects Shape Id Shape Number of Graphical Coverage to overall properties objects objects Shape-1 Rect: Rx 0 4 X axis, Y axis, 100 Ry 0 X axis line, background 687347 O -17o Fig. 8 is a simplified version of Fig. 3 illustrating the graphical objects expressed in Table 1 and Table 2. A caption object 810 has two styles extracted, identified as Style-1 for colour and Style-2 for font style. Style-3 is the colour of the background t object 860 that occupies the most area (about 60%) of the template. Style-4 is shared by 2 rectangles, one for Y axis 820 and another for the bottom part of X axis 840. N1 is the colour for the top line of the X axis 850. Style-6 and Style-7 specify the colour and Sfont for the 11 text objects of Y axis scales marked as 0 to 100 in Fig. 3 and a percentage sign symbol The total area covered by these text objects is labelled by 830. Note that only the colour style can have a list of adjacent colours specified in Table 1.

Similarly, a style statistic table 630a and a shape statistic table 630b for the data graphical objects can be constructed in Table 3 and Table 4, respectively.

Table 3: Style statistic table for the data graphical objects Style Id Styles Num. Graphical Coverage Adjacent of objects to overall colour list objects Style-d-1 Colour =5678DC 4 3 rectangles at the 25 D6D6D7 bottom, 1 square at the top Style-d-2 Colour 8C102D 4 3 rectangles at the 27 D6D6D7 top, 1 square at the top Style -d-3 Colour 3 3 shades 10 D6D6D7 =LinearGradient Style-d-4 Colour =FFFFFF 3 Text for X axis 15 8C102D scales 687347 0 -18- Font: size=1 6 3 Text for X axis weight=bold scales Style -d-6 Colour =565656 2 2 legends at the 8 D6D6D7 top O Style -d-7 Font: size=1 6 2 2 legends at the 8 cn weight=bold top Table 4: Shape statistic table for the non-data graphical objects Shape Id Shape Number of Graphical objects Coverage to overall properties objects ove Shape-d-1 Rect: Rx 0 8 6 bars and 2 legends 52 Ry 0 Shape-d-2 Rect: skewed 3 3 shades for the bars angle= 4 Once these four tables 630 are constructed by the process 620, the process 640 can proceed to obtain the dominant styles including dominant shapes. Depending on the statistics of styles and shapes extracted, the dominant styles are obtained based on the styles shared by the graphical objects covering most of the area of the template.

Examples of the dominant styles are: 1. Dominant colours; 2. Dominant fonts; 3. Dominant shapes; 4. Dominant borders.

Fig. 9 is a flowchart executed by the process 640. The process starts at step 910 where styles are grouped by their names, all colour styles are under one group and all 687347 -19font styles are under a different group. The shape table is considered as one style group.

These groups of styles are denoted as {Si}.

At step 915, a style group Si is selected. For example, the colour style is first t chosen. Then, step 920 operates to initialise a dominant style list DSLi corresponding to the selected style group, DSLi is a dominant colour list. Each element in the N dominant style list is a reference to a row in either style or shape statistic table. At step 925, the maximum number of elements, MaxNumStyles, in the dominant style list is initialised. This number is like a configuration setting. For example, if three dominant colours are considered, MaxNumStyles is set to 3.

At step 935, a check is made to see whether the dominant style list DSLi is a dominant shape list. If not, the styles whose graphical objects cover the most area are identified from the style statistic tables 630a at step 940. If the style statistics table 630a for the non-data graphical objects Table 1) does not have MaxNumStyles styles, the style statistic table 630a for the data graphical objects Table 3) is used to find the dominant styles.

Step 950 creates a number of reference entries in the dominant style list and each reference entry points to the corresponding row in the style statistic table. A decision step 955 then checks whether more dominant style lists, dominant font list, are to be processed. If so, step 915 is repeated. Otherwise, the process terminates at step 980.

At decision step 935, if the dominant style list is a shape list, step 960 finds MaxNumStyles dominant shapes from the shape statistic tables for data and non-data graphical objects. Then step 970 creates the same number of reference entries in the dominant shape list. Each reference entry points to a shape in the shape statistic table 630b.

Fig. 11 illustrates the indirect references from the dominant style lists 1120 1140 687347 o (shown as 650 in Fig. to the style and shape statistic tables 1110 and 1150 (shown as 630 in Fig. The most dominant colour, font and shape are at the top of each of the corresponding dominant style lists DC 1120, DF 1130 and DS 1140, respectively. Each t element in the dominant colour list 1120 and dominant font list 1130 points to a row in 5~5 the style statistic table 1110 while each element in the dominant shape list 1140 points to Cc- N a row in the shape statistic table 1150. In the example of the bar chart template in Fig. 3, the most dominant colour is defined by the style-3 for the background 380 of the template. The second dominant colour is defined by Style-4 for XY axes 340 and 360.

The most used shape is a sharp cornered rectangle specified by Shape-i.

The process 660 either generates a new style family or integrates the dominant style lists 650 into an existing style family. Before the process 660 is described, the style family shall be defined.

Fig. 12 illustrates an example of the style family based on three templates 520 526 (see Figs. In a preferred implementation, a style family 1200 includes: 1. aprimary style set 1212; 2. a secondary style set 1214; and 3. a set of style examples 1216.

The primary style set 1212 preferably includes the following data: 1. a set of dominant colour lists 1220 denoted as {FamilyDCi} where i is the index to a dominant colour list in the set; 2. a set of dominant font lists 1230 denoted as {FamilyDFj} where j is the index to a dominant font list in the set; and 3. a set of most used shape lists 1240 denoted as {FamilyDSk} where k is the index to a dominant shape list in the set.

687347 0 -21- Each set may contain a different number of dominant lists.

The style table 1245 and shape table 1248 in the primary style set 1212 are constructed using the dominant style lists 650. However, only styles with coverage statistics and adjacent colour lists are copied over. For example, the styles in the primary 5 style set for Fig. 3 is shown in Table 5 below: C1 Fig. 5: Style table in primary style set 1 Style Id Styles Coverage to Adjacent colour list overall Style -3 Colour =D6D6D7 Style -4 Colour =3A3A38 20 D6D6D7, 8C102D Style -6 Colour =FFFFFF 10 3A3A38 Style -2 Font: size=14, weight =bold Style-7 Font: size=12, weight=plain The secondary style set 1214 provides supplementary styles for a template. The set 1214 is constructed from the remainder of the style and shape statistic tables 630 that are not considered as dominant styles. The secondary style set 1214 can include the following styles: 1. a list of available colours 1250; 2. a list of available fonts 1260; 3. a list of shapes 1270; and 4. a list of colour pairs that are generated from the adjacent colours in the style tables 1280.

687347 S- 22- The style example set 1216 contains a list of dominant style reference entries 1291 to 1293 to the primary style set 1212. Each reference entry, (RefFamilyDCi, RefFamilyDFj, RefFamilyDSk) refers to corresponding dominant style lists FamilyDCi, I FamilyDFj, FamilyDSk in the primary style set 1212.

O

5 The process 660 of "adapt styles" is now described. When a template is designed I by the graphic designer for a family of templates, the styles of the template must be Saccepted by the designer as a part of the template family, according to the particular design philosophy. If another template is designed by a different graphic designer for the same family, its styles may be somewhat different, a new colour may be added.

However, the new colour should be incorporated into the rest of the style family. Thus, one way of adapting the styles from one template is accepting its styles and accumulating them into the style family 680.

Fig. 13A is a flow chart of a sub-process 1300 executed by the process 660 in order to generate a style family from a first template. The sub-process 1300 starts at step 1302 to copy the dominant style list to the primary style set of the style family. For example, the dominant colour list DC 1120, the dominant font list DF 1130 and the dominant shape list DS 1140 are copied to the corresponding dominant style set {FamilyDCi}, {FamilyDFi}, {FamilyDSi}, respectively. The brackets represents a set of dominant style lists, {FamilyDCi} is a set consisting of one dominant colour list DC 1120 already in the style family. Then, at step 1303 the styles and shapes defined in the style and shape statistic tables 630 (1110,1150), and referenced by the dominant style lists (DC, DF, DS) 1120-1140 are copied to the style table 1245 and shape table 1248 in the primary style set together with coverage percentages and the list of adjacent colours.

Next, step 1304 follows to create a reference entry, (RefFamilyDCi, RefFamilyDF 1 RefFamilyDSi) that refers to the three dominant style lists, FamilyDCi, 687347 -23o FamilyDF1 and FamilyDS1. The reference entry is added to the style example set 1216 of the style family.

Finally, step 1306 adds the rest of the styles and shapes into the secondary style t set. The sub-process 1300 then terminates at step 1308.

Fig. 13B depicts the flowchart of a further sub-process 1399 executed by (Ni N, process 660 when the style family is not empty. The sub-process 1399 starts at step 1310 which retrieves the dominant style lists 630, (DC, DF, DS) of the template 610 and the corresponding dominant colour set {FamilyDCi} 1220 dominant font set {FamilyDFj} 1230 and dominant shape set {FamilyDSk} 1240 from the primary style set in the existing style family.

Then step 1320 follows to initialise a reference entry RE. The reference entry RE includes references to the dominant style list in the primary style set. Step 1330 continues to get a dominant list from the dominant styles 630 (DC, DF, DS) of the template 610 and this dominant style list is assigned to a variable S. A corresponding dominant style set in the primary style set is also obtained as {FamilySi}. For example, if the dominant colour list DC 1120 is assigned to S, the dominant colour set {FamilyDCi} is obtained and assigned to {FamilySi}.

A check is made at step 1335 to determine whether the dominant list S is close to any of the dominant style lists in {FamilySi}. If S is the dominant colour list DC 1120, a colour comparison method, forming part of step 1335, is invoked. The first colour in the dominant colour list DC 1120 is compared with the first colour in the FamilyDCi, and then second colours are compared, and so on. Two dominant colour lists 1120 are close if three colours in one dominant colour list 1120 are visually close to three colours in another dominant colour list in FamilyDCi. Two colours are considered visually close if they have the same hue and saturation but are slightly different in brightness. A threshold 687347 -24can be defined to determine the closeness of two colours. Other methods of colour comparison can be implemented in the colour comparison method of step 1335.

If S is the dominant font list DF 1130, a font comparison method is used in t step 1335. This compares fonts in S to the fonts in DF 1130. In a simple case, two fonts 5 are considered close, if the difference between two font sizes is less than 1 point, two font N, styles are the same both are italic or both are normal) and two font weights are same (bold, no-bold). More sophisticated font comparison methods can be defined and built into the process 660.

If S is the dominant shape list DS 1140, a shape comparison method is used in step 1335. When comparing the shapes for dominant shapes, not only is the shape name used, but also the shape's properties, such as radius for the rectangle corners, are required.

Shape position and size are not used for the comparison as these are not inherent in the concept of style, but rather are specific implementations of a style. More complicated shape comparison methods can be built into the process 660 for different kinds of shapes and their properties.

If a dominant style list S has a close match to another dominant style list FamilySi, a reference RefFamilySi to the dominant style FamilySi is appended to a reference element RE at step 1340. For example, RE contains (RefFamilyDC1, RefFamilyDF 1 RefFamilyDS 1 referring to the dominant style lists FamilyDCI, FamilyDFl and FamilyDS1.

At step 1345, a check is made to see whether there are any unchecked dominant lists from (DC, DF, DS). If so, the execution moves to step 1330. Otherwise, the reference entry RE is added to the style example set 1216 at step 1385 and the remaining styles are added to the secondary style set at step 1390. Then the process terminates at step 1395.

687347 o Returning to decision step 1335, if there is no close match between a dominant d style list, DC and any dominant style list in {FamilySi}, {FamilyDCi}, there is a need for assistance from a built-in knowledge base or a style adviser 670, seen in Fig. 6.

t Step 1350 follows to invoke the style adviser 670 to check each style in the dominant 0 5 style list S of the template with the styles in the style family. The style adviser 670 can

CC-

N include a colour adviser, design principles, and even predetermined rules of thumb. For example, from the dominant colours in the style family, a colour scheme can be determined. When a new colour from a template to be integrated into the family is checked, the colour adviser in the style adviser 670 operative in step 1350 determines whether this colour belongs to the same colour scheme. Similarly, if the fonts in a style family are all italics and a template to be integrated into the family has a non-italic font, the style adviser 670 can detect the inconsistency.

If any inconsistency in a dominant style list S is found by the style adviser 670 at step 1350, a dialog box is displayed via the GUI on the display 114 asking the graphic designer whether this inconsistency should be accepted at step 1355. If it is accepted by the graphic designer, the dominant style list S in question is assigned to FamilySi+l and that family is appended to the dominant style set {FamilySi} at step 1380. Then the execution of the sub-process 1399 goes back to step 1340.

If the graphic designer does not accept the dominant style list S, step 1360 continues to propose a style list using the knowledge of the style adviser 670.

However, the graphic designer may not like the suggested styles from the style adviser 670. The graphic designer accepts or rejects the styles at step 1365. If accepted, the advised style is assigned to S at step 1370 and then the execution continues to step 1380. If it is not accepted, the graphic designer needs to create a style or a style list and assign that created style to S at step 1375. Both style lists suggested by the style 687347 -26- Sadviser 670 and created by the graphic designer are integrated into the style family {FamilySi}at step 1380. However, any style rejected by the graphic designer is built into the knowledge base in the style adviser 670 as a bad example and should not be suggested t again.

Returning to Fig. 6, the process 690 of creating a virtual template 695 may now be described. A virtual template is initially copied from the template 610 and then modified to include references to the styles in the style family 680. For example, the graphical objects that use the first dominant colour are expressed by DCI. Only the graphical objects that have dominant styles are modified using indirect references. The rest of the styles are not changed.

Fig. 14 depicts a flowchart of the process 690 of creating a virtual template 695 and Fig 15 shows an example of a virtual template in SVG for Fig. 3 with partial data.

The process 690 begins at step 1402 to copy the template 610 to a virtual template 695.

Then step 1405 follows to update the virtual template with a new name space 1510 to use the style family 680. Step 1410 gets a dominant list from (DC, DF, DS) and assigns it to a variable DL. It sets an index I, initially to 0 to start getting the elements of the dominant list DL. Step 1420 gets a reference element R from DL. The reference points to a row in the style or shape tables 1245 and 1248. Then this row is extracted in step 1430.

At step 1440, an element E with a dominant style name is created and inserted into the virtual template 695 in SVG. Each dominant style S referenced by R, with coverage percentage, if any, is also inserted into the element E. For example, elements DC_1, DC 2 and DC 3 for the first three dominant colours are created as 1520, 1530 and 1540 in Fig. 15. The graphical objects using DC_1 cover 60% of the template space.

Each row in the style table or shape table also contains a list of graphical objects that use the dominant style specified in the same row. Step 1450 follows to get this list as 687347 -27- Sa graphical object list OL. A check is made at step 1455 to determine if the graphical object list OL is empty. If not, a graphical object is obtained from the object list OL at step 1460. At decision step 1485, a check is made to determine whether DL is a dominant t shape list. If not, the graphical object's corresponding style is replaced with a reference to the element E at step 1470. For example, as seen in Fig. 15 the SVG code line 1550 (Ni N, defines a rectangle object and its fill colour is changed to #DC_1. If DL is a shape list, a property 'MostUsedShape' with a value referencing the element E, #DS_I, is inserted at step 1480. Following each of steps 1470 and 1480, control returns to the decision step 1455.

When all graphical objects in OL are changed to refer to the dominant styles, and thus the OL is empty, a check is made at step 1465 to see whether there are more styles in the dominant list DL. If so, the process 690 goes back to step 1420 to get next reference R. Otherwise, a check on the unprocessed doriinant list is made at step 1475.

If there is another dominant list, dominant font list DF, that list is assigned to DL at step 1410. Otherwise, the process 690 terminates at step 1490.

The virtual template cannot be used alone because of the indirect references to styles. However, the virtual template can be used to generate templates.

As described in Fig. 5B, the template generator 240 uses a style family 546 and one of the virtual templates 550, 552 or 554 to create a new template 560, 562 or 564.

Some of the graphical objects in the virtual templates have references to the dominant styles in the style family. Therefore, the new template can be produced by replacing the references with the specific styles defined in the style family.

Fig. 17 is a flowchart for a process 1700 for generating a template performed by the template generator 240. The process 1700 starts at step 1705 to select a virtual template 1500 and a style family 1200 or 680. Then step 1710 follows to find a reference 687347 -28- Selement defined at the top of the virtual template. For example, line 1520 in Fig. 15 is a reference element for the first dominant colour and is represented as <DC_1 id="FirstDominantColour" t The name of the element is then decomposed into the name of the dominant style, 0\ 5 DSN, and the index to the list, I at step 1720. In the above example, DC_1 is separated as DC for dominant colour, DSN is equal to DC, and 1 is for the first dominant colour.

Some reference elements contain the coverage percentage of the graphical objects in the template. In this case, the coverage is assigned to a statistic variable St at step 1730. In the above example, 60% is the coverage percentage for the first dominant colour in the current virtual template.

At step 1740 the dominant style name DSN is used to get a corresponding dominant style set {FamilyDSi} 1220, 1230 or 1240 in the primary style set 1212 in the style family 1200 or 680. Then step 1750 gets a dominant style list, FamilySi, from {FamilySi}. At decision step 1755, a check is made to determine any unprocessed reference in the dominant style list FamilySi, for example, any dominant colours in the dominant colour list 1120 not yet checked. If so, step 1760 continues to get a reference R by index I and then extracts a style S and a corresponding statistic St from the row referred to by R in the style or shape tables 1245 or 1248 in the primary style set 1212.

For example, colour D6D6D7 and 60% of coverage in the first row of Table 5 are extracted. Step 1765 determines if the coverage from the virtual template is close to the coverage in the referenced row in the style family 1200. If they are, the extracted style is used to replace the reference DC_I in every graphical object in the virtual template at step 1780. For example, #DC_1 in 1550 is replaced by colour D6D6D7. Decision step 1785 checks for any unprocessed reference element in the virtual template 1500. If so, the process 1700 repeats at step 1710. Otherwise, this process 1700 terminates at 687347 -29o step 1795.

Decision step 1775 follows when there is no match between the statistics from the virtual template and the statistics from the style or shape table 1245 or 1248 at step 1765.

I Step 1775 searches for the two statistics which are closest. For example, a template with 5 three dominant colours covering 40%, 10% and 5% of the template space, and may be C associated with the dominant colour list to the style family that covers 60%, 8% and 3%.

As more than one comparison can be done between the statistic of the virtual templates and other statistics of the {FamilySi}, a closest match between two statistics can be determined and is remembered as CS at step 1790. Then, step 1755 is repeated. If the comparison of two statistics at step 1775 is not the closest, step 1755 is repeated.

When there are no more references to process at decision step 1755, another decision step 1745 is checked to see whether there is any unprocessed FamilySi from the {FamilySi}. If so, step 1750 is repeated. Otherwise, the closest dominant style list CS is assigned to the style S at step 1770 and then S is used to replace the style reference in the graphical objects in the virtual template at step 1780.

The generated template using the style family and the virtual template contains all specific styles, no style reference is used. It can be viewed independently, as any template 610.

As the templates in a family can have different formats, graphical objects can be labelled arbitrary, the styles referred to by the graphical objects in one template may not be easily mapped or applied to the graphical objects in another template. However, the dominant styles can always be obtained by analysing the template. Thus, referencing to a specific style via references to the dominant styles provides a platform to share styles among templates in different formats.

687347 It follows from the above description that a representation of data can be derived from existing representations, formats and styles. Fig. 10 summarises a general approach 1000 in which a number of existing representations 1002 are analysed in terms of shape and style. Shape analysis reveals a skeletal structure of each representation in terms of those graphical objects that make up each representation. From this example, C those skeletal structures 1006 are seen to be a bar chart, a pie chart, a graph and a table.

Style analysis of the representations 1002 identifies the various features 1004 of each representation (eg. font type, font size, line type, fill style, colours etc.) which are then formed into dominant families 1008 according to a statistical occurrence of a style across the respective existing representation 1002. For example, one family may be formed by Arial font at 14 point size, with solid lines of 8 point size and solid colours including cyan magenta yellow and black Once the various shapes and the style families have been distilled from the example representations 1002, selected examples of each may be combined to form a template 1010 to which new data 1012 is applied thus revealing a new representation 1014 based on the new data but incorporating the selected skeletal shape and the selected style family.

The approach of Fig. 10 may be extend in the fashion shown in Fig. 16A. In this implementation 1600, corresponding steps and features of Fig. 10 are having corresponding functions are represented by the same reference numerals. However, from Fig. 16A, it will be seen that, rather than analyse the styles of the existing representations 1002, a library 1602 of dominant style families is provided from which the user may select to combine with a selected one of the analyses shapes, to form a template 1606 based on the selection and the performed analysis. The template 1606 is combined with new data 1608 to then give a new representation 1604 which is seen to be different from the representation 1012 of Fig. 10 in respect of font type and size.

687347 -31- Fig. 16B illustrates a complementary approach 1610 to that of Fig. 16A where the shapes are selected from a library 1614 and the styles are analysed from the existing representations 1002, in this example revealing a new representation 1612 that differs from each of the representations 1012 and 1604.

Once a template 1606 has been created, by a graphical designer, or generated from N virtual template and a style family using the approaches of the present disclosure, the template can be used to present new data that is suitable for the template format. For example, a table of numerical data can be presented with a bar chart template. Although the template has sample data for the data graphical objects, at least one of these data graphical objects will be used to present the new data. More data graphical objects can also be created.

In the example of the bar chart template of Fig. 3, although there are two series of bars, in which each series of bars includes three bars (eg. 312, 314, 316), a new bar chart presentation may be different from this sample data (eg. four series of bars, with each series having five bars). However, the visual effects remain the same as established by the template.

Industrial Applicability The arrangements described are applicable to the computer and data processing industries and particularly where a data set is intended to be represented according to a desired style, which may be used for other data sets. This particularly useful for graphical representations of data sourced from heterogeneous data sources. Although the present invention is described by example for computer graphic design, it can be applied to many areas involving styles such as interior design and fashion design and computerised processes for their creation, development and modification. Using dominant styles across different design domains can provide an aesthetically pleasing result.

687347 -32- The foregoing describes only some embodiments of the present invention, and modifications and/or changes can be made thereto without departing from the scope and spirit of the invention, the embodiments being illustrative and not restrictive.

S(Australia Only) In the context of this specification, the word "comprising" means "including principally but not necessarily solely" or "having" or "including", and not C "consisting only of'. Variations of the word "comprising", such as "comprise" and S"comprises" have correspondingly varied meanings.

687347

Claims (24)

1. A method of generating a style family and virtual template from existing t templates, the existing templates being arranged in different formats, said method comprising the steps of: S(a) extracting styles and shapes from at least one said existing template into style and shape statistic tables by analysing each graphical object in the said one existing template; obtaining one or more dominant styles from said style and shape statistic tables, said dominant style being a style shared by one or more graphical objects that cover prominent areas of a space of said one existing template; adapting said dominant styles from the style and shape statistic tables to thereby generate said style family; and forming a virtual template based on said one existing template, said virtual template having at least one graphical object referring to styles in said style family.

2. A method according to claim 1 wherein steps to are performed for each said existing template using common style and shape statistic tables to thereby accumulate styles in said style family, step being performed with a selected one of said existing templates.

3. A method according to claim 1, wherein step comprises the sub-steps of: (ca) accumulating said styles and shapes with corresponding usage statistics for said dominant styles to form a primary style set in said style family,; 687347 -34- o (cb) adding styles that are non-dominant styles into a secondary style set for available styles; and (cc) adding a reference entry to a style example set in said style family, said I reference entry containing references to the dominant styles in said primary style set. 1

4. A method according to claim 3 wherein said usage statistics comprise a coverage percentage in said space of said existing template.

A method according to claim 1 or 3, wherein step further comprises further integrating a style adviser to check consistency between said dominant styles from said template and a remainder of the style family.

6. A method of generating a template from a style family and a virtual template formed according to the method of any one of claims 1 to 5, said method comprising the steps of: obtaining dominant style names from said virtual template; comparing said dominant style names with the dominant styles in the style family; and if the style names match the dominant styles, modifying the graphical objects in said virtual template referring to the dominant styles with a style specified in the style family.

7. A method according to claim 6 further comprising the step of: if the style names do not match the dominant styles, modifying the graphical objects of said virtual template with styles obtained from a style adviser. 687347 U

8. A method of forming a representation of present data, said method comprising: analysing shapes of graphical objects present in at least one existing Srepresentation of data to form at least one template for the representation of data; analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; selecting at least one said template and applying one said style family to graphical objects thereof to form a stylized template; and applying said present data to said stylized template to thereby form a representation of said present data.

9. A method of forming a representation of present data, said method comprising the steps of: analysing shapes of graphical objects present in at least one existing representation of data to form at least one template for the representation of data; selecting, from a library of style families of graphical objects, a style family desired to be used in representing said present data; selecting at least one said template and applying said selected style family to graphical objects thereof to form a stylized template; and applying said present data to said stylized template to thereby form a representation of said present data.

A method of forming a representation of present data, said method comprising: 687347 C -36- analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; selecting, from a library of templates of graphical objects structured for the representation of data, a template for the representation of said present data; N selecting at least one said style family and applying said selected style Ofamily to graphical objects of said selected template to form a stylized template; and applying said present data to said stylized template to thereby form a representation of said present data.

11. A computer readable medium having a computer program recorded thereon and adapted to make a computer execute a procedure to generate a style family and virtual template from existing templates, the existing templates being arranged in different formats, said program comprising: code for extracting styles and shapes from at least one said existing template into style and shape statistic tables by analysing each graphical object in the said one existing template; code for obtaining one or more dominant styles from said style and shape statistic tables, said dominant style being a style shared by one or more graphical objects that cover prominent areas of a space of said one existing template; code for adapting said dominant styles from the style and shape statistic tables to thereby generate said style family; and code for forming a virtual template based on said one existing template, said virtual template having at least one graphical object referring to styles in said style family. 687347 -37-

12. A computer readable medium according to claim 11 wherein code for extracting, adapting and obtaining are executable for each said existing template using common style and shape statistic tables to thereby accumulate styles in said style family, the code for forming being executable with a selected one of said existing templates. 1

13. A computer readable medium according to claim 11, wherein said code for Oobtaining comprises: sub-code for accumulating said styles and shapes with corresponding usage statistics for said dominant styles to form a primary style set in said style family,; sub-code for adding styles that are non-dominant styles into a secondary style set for available styles; and sub-code for adding a reference entry to a style example set in said style family, said reference entry containing references to the dominant styles in said primary style set.

14. A computer readable medium according to claim 13 wherein said usage statistics comprise a coverage percentage in said space of said existing template.

A computer readable medium according to claim 11 or 13, wherein said code for obtaining further comprises code for further integrating a style adviser to check consistency between said dominant styles from said template and a remainder of the style family.

16. A computer readable medium having a computer program recorded thereon and adapted to make a computer execute a procedure to generate a template from a style 687347 C -38- ofamily and a virtual template formed by execution of the program of any one of claims 11 to 15, said program comprising: code for obtaining dominant style names from said virtual template; code for comparing said dominant style names with the dominant styles in the style family; and N code for modifying, if the style names match the dominant styles, the graphical Oobjects in said virtual template referring to the dominant styles with a style specified in the style family.

17. A computer readable medium according to claim 16 further comprising: code for modifying, if the style names do not match the dominant styles, the graphical objects of said virtual template with styles obtained from a style adviser.

18. A computer readable medium having a computer program recorded thereon and adapted to make a computer execute a procedure to form a representation of present data, said program comprising: code for analysing shapes of graphical objects present in at least one existing representation of data to form at least one template for the representation of data; code for analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; code for selecting at least one said template and applying one said style family to graphical objects thereof to form a stylized template; and code for applying said present data to said stylized template to thereby form a representation of said present data. 687347 -39-

19. A computer readable medium having a computer program recorded thereon and adapted to make a computer execute a procedure to form a representation of present data, said program comprising: O code for analysing shapes of graphical objects present in at least one existing representation of data to form at least one template for the representation of data; Scode for selecting, from a library of style families of graphical objects, a style family desired to be used in representing said present data; code for selecting at least one said template and applying said selected style family to graphical objects thereof to form a stylized template; and code for applying said present data to said stylized template to thereby form a representation of said present data.

A computer readable medium having a computer program recorded thereon and adapted to make a computer execute a procedure to form a representation of present data, said program comprising: code for analysing styles of graphical objects present in at least one existing representation of data to form style families applicable to graphical objects in the representation of data; code for selecting, from a library of templates of graphical objects structured for the representation of data, a template for the representation of said present data; code for selecting at least one said style family and applying said selected style family to graphical objects of said selected template to form a stylized template; and code for applying said present data to said stylized template to thereby form a representation of said present data. 687347 U a,)

21. Computer apparatus for generating a style family and virtual template from existing templates, the existing templates being arranged in different formats, said t' apparatus comprising: means for extracting styles and shapes from at least one said existing template into N, style and shape statistic tables by analysing each graphical object in the said one existing template; means for obtaining one or more dominant styles from said style and shape statistic tables, said dominant style being a style shared by one or more graphical objects that cover prominent areas of a space of said one existing template; means for adapting said dominant styles from the style and shape statistic tables to thereby generate said style family; and means for forming a virtual template based on said one existing template, said virtual template having at least one graphical object referring to styles in said style family.

22. A method of generating a style family and virtual template from existing templates, said method being substantially as described herein with reference to any one of the embodiments as that embodiment is illustrated in the drawings.

23. A template formed according to the method of any one of claims 1 to 10 or 22.

24. A computer readable medium having a program recorded thereon and adapted to make a computer execute a procedure to perform the method according to claim 22. 687347 S-41- Computer apparatus for generating a style family and virtual template from d) existing templates, said method being substantially as described herein with reference to any one of the embodiments as that embodiment is illustrated in the drawings. Dated this FOURTEENTH Day of DECEMBER 2004 CN CANON KABUSHIKI KAISHA 0Patent Attorneys for the Applicant Spruson&Ferguson 687347