Methods for interactively defining transforms and for generating queries by manipulating existing query data

• US 20040015783A1
• Filed: 06/20/2003
• Published: 01/22/2004
• Est. Priority Date: 06/20/2002
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Wu, Jing, Lennon, Alison Joan, Van Doan, Khanh Phi, Chan, Sammy, Qian, Lena, Yap, Sue-Ken, Worrill, Joshua Brendon, Edgoose, Timothy Charles
• Granted Patent: US 7,574,652 B2
• Time in Patent Office: Days
• US Class Current: 715/523
• CPC Class Codes: G06F 16/242 Query formulation
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/258 Data format conversion from or to a database

Abstract

Methods, programs (120) and apparatus (100) are disclosed for accessing heterogenous data sources (150-153). Specifically, access maybe obtained by a method of interactively defining transformations for use in mapping at least one source data component from at least one data source to a target data component, includes selecting (500) at least one source data component. For each selected source data component, a corresponding list of representative data examples is generated (510). A resulting list of representative data examples for the target data component is compiled from the lists of representative data examples of the selected source data components, the resulting list of data examples being representative of the target data component. An example (514) from the resulting list of examples representative of the target data component is then modified (516), the example being modified as a single textual entity. Finally, at least one transformation to be applied to at least one source data component is inferred (518) from the modified example. Alternatively, access may be obtained by a method of forming a query from manipulations of existing query data in a graphical user interface (1200). The existing query data typically includes one or more views of data (5405), with each data view being a visualisation of a data structure and associated with a visualisation type and a query, the query defining how to obtain data for the data structure of the data view. This involves selecting a data component from a first existing data view. The selected data component is then indicated to be added to a second data view (5420), where a target location of the selected data component in the second data view indicates a role of the selected data component in the second data view, the role being dependent on the visualisation type associated with the second data view. A query for the second data view is then created (5420), and which specifies how to obtain data for the data structure for the second data view. The query is formed using the indicated data component and an interpretation (5425) of its corresponding role in the second data view. When this is accomplished the selected data component is added to the second data view (5435). Other methods associated access are also dislcosed.

Claims

1. A method of interactively defining transformations for use in mapping at least one source data component from at least one data source to a target data component, said method comprising the steps of:

(a) selecting at least one said source data component;

,

(b) for each said selected source data component, generating a corresponding list of representative data examples;

(c) compiling a resulting list of representative data examples for said target data component from said lists of representative data examples of said selected source data components, said resulting list of data examples being representative of said target data component;

(d) modifying an example from said resulting list of examples representative of said target data component, said example being modified as a single textual entity; and

(e) inferring from said modified example, at least one transformation to be applied to at least one source data component.

2. A method according to claim 1, wherein step (b) is performed upon each selection in step (a), and step (c) comprises the sub-steps of:

(ca) initialising said resulting list of representative data examples for said target data component to said list generated in step (b) corresponding to a first selected data source component; and

where at least one subsequent said data source component was selected in step (a),

(cb) concatenating each example of said resulting list with a corresponding example in a list generated for said subsequent data source component.

3. A method according to claim 2, wherein step (a) comprises performing an ordered selection of said source data components, and said modifying comprises selecting and editing said example.

4. A method according to claim 1, comprising the further step of:

(f) updating said resulting list of representative examples using said inferred one or more transformations and said data components obtained from said one or more data sources.

5. A method according to claim 1, comprising the further step of:

(g) representing the resulting inferred transformation using a stylesheet.

6. A method according to claim 1 where said selection of source data components comprises at least one of:

(i) a selection of indicia representing the source data components; and

(ii) a selection of structured data, said structured data comprising a list of representative data examples for said source data component.

7. A method according to claim 1 wherein said inferring implements a determination by example to derive said transformation.

8. A method according to claim 7, wherein said modifying comprises an interpretation of a user entered demonstration of a transformation, which structurally transforms the data, and said inferring considers only that part of the example that has been changed by the said modifying step.

9. A method according to claim 1, wherein said transformation is a unary transform, said unary transform having a single operand, and said inferring considers solutions in which a sequence of unary transforms are performed, with the operand of each step in the transform sequence being the result of the previous step.

10. A method according to claim 1, wherein said transformation is a binary transform involving sequential pairs of data operands.

11. A method according to claim 1, wherein said transformation is an n-ary operation involving at least two data operands, and said inferring considers solutions in which a sequence of binary transforms are performed with the first operand of each step in the transform sequence being the result of the previous step.

12. A method according to claim 9 wherein said unary transform is selected from the group consisting of:

(i) a string containing the first n words of the operand;

(ii) a string containing n words starting from start word;

(iii) a string containing first n characters of the operand;

(iv) a string representing operand converted to upper case;

(v) a string representing operand converted to lower case;

(vi) a string in which the first characters of all non-conjunction words in the operand are capitalised;

(vii) a string in which the first characters of all words in the operand are capitalised;

(viii) a string in which the operand is translated to the language specified by the XML language;

(ix) a string in which all punctuation in the operand is removed;

(x) a string in which the string text is inserted at position n in the operand;

(xi) a string in which all the conjunctions in the operand have been removed;

(xii) a number if the operand can be parsed as a number;

(xiii) a string representing the operand;

(xiv) a value which is the negation of the operand;

(xv) a value of the operand as an integer; and

(xvi) a value of the operand as double precision number

13. A method according to claim 10 wherein said binary transforms include mathematical functions selected from the group consisting of addition, subtraction, division and multiplication.

14. A method according to claim 11, wherein said n-ary transforms include operations selected from the group consisting of min, max and sum.

15. A method according to claim 1, wherein said inferring comprises interpreting a user entered demonstration of a transformation, which defines a style in which at least a part of the transformed data is to be presented, and said demonstration includes at least one of:

(i) a style applied to the entirety of the result of the transformation; and

(ii) a style applied to data arising from at least one said source component involved in the mapping.

16. A method according to claim 15 wherein said demonstration of style asserts at least one of:

(i) that the data be displayed using a style characteristic selected from the group consisting of bold, italic, underline, superscript, and subscript; and

(ii) that the data be displayed using at least one of a specified font type, font size and colour.

17. A method according to claim 1 wherein said target data component is stored as a named component of a unique namespace, and said target data component represents a component of a schema view over the at least one selected data source.

18. A method according to claim 2 wherein generation of said target data component is specified using the W3C Extensible Stylesheet Language Transformation Recommendation,

19. A method according to claim 18 wherein a stylesheet resulting from said generation is used to re-purpose components of the at least one selected data source for presentation.

20. A method according to claim 1 wherein said target data component of said mapping comprises a dimension consistent with a dimension of the corresponding said source data component.

21. A method according to claim 1 wherein said mapping defines units of said dimension for said target view component and data associated with said source data components of the mapping are converted into said defined units.

22. A method according to claim 1, further comprising extending a set of mappings, said set of mappings acting to provide a unified representation of a collection of heterogeneous data sources and comprising at least one said mapping, said target data component being a constituent part of said unified representation, said extending comprising the steps of:

(1) receiving a data view, said data view having been created for the purposes of exchanging a view of data derived from one or more of said heterogeneous data sources and containing at least one target data component, said target data component resulting from the application of a mapping;

(2) importing, as a constituent part of said set of mappings, mappings associated with said received data view, said importing creating an extended set of mappings; and

(3) creating at least one new data view using said extended set of mappings.

23. A method according to claim 1, further comprising visualising a collection of heterogeneous data sources using a set of mappings, said set of mappings acting to provide a unified representation of the said collection and comprising at least one said mapping, said target data component being a constituent part of said unified representation, said visualising comprising the steps of:

(1) selecting a set of mappings to use for the visualisation;

(2) selecting at least one data source to be visualised;

(3) identifying, for each selected data source, a schema definition representative of said selected data source, and using said definition to identify all child definitions for said data source, to thereby form a schema view for the selected data sources;

(4) generating a tree representation for said schema view;

(5) applying each of said mappings of said mapping set to said schema view, each application of each said mapping comprising the steps of;

(5a) identifying, for said mapping, said one or more source data components;

(5b) if all said identified source data components exist in said schema view, then performing steps (5ba), (5bb) and (5bc) as follows;

(5ba) identifying the target data component of said mapping;

(5bb) creating a visual representation of said target data component;

(5bc) updating said schema view by inserting the created visual representation of said target data component into said schema view such that the hierarchical context of the target data component in the schema view is the same as a maximum common hierarchical context of the source view components of the said mapping; and

(6) visually presenting the updated schema view.

24. A computer readable medium, having a program recorded thereon, where the program is configured to make a computer execute a procedure to interactively define transformations for use in mapping at least one source data component from at least one data source to a target data component, said program comprising:

code providing for selection of at least one said source data component;

code for generating a list of representative data examples for each corresponding said selected source data component;

code for compiling a resulting list of representative data examples for said target data component from said lists of representative data examples of said selected source data components, said resulting list of data examples being representative of said target data component;

code for modifying an example from said resulting list of examples representative of said target data component, said example being modified as a single textual entity; and

code for inferring from said modified example, at least one transformation to be applied to at least one source data component.

25. Computer apparatus for interactively defining transformations for use in mapping at least one source data component from at least one data source to a target data component, said apparatus comprising:

means for selecting at least one said source data component from at least one heterogenous data source;

means for generating, upon each selection of one said data source component, a list of representative data examples for each corresponding said selected source data component;

means for compiling a resulting list of representative data examples for said target data component from said lists of representative data examples of said selected source data components, said resulting list of data examples being representative of said target data component;

means for modifying an example from said resulting list of examples representative of said target data component, said example being modified as a single textual entity; and

means for inferring from said modified example, at least one transformation to be applied to at least one source data component.

26. A method for forming a query from manipulations of existing query data in a graphical user interface, said existing query data comprising one or more views of data, each said data view being a visualisation of a data structure and associated with a visualisation type and a query, said query defining how to obtain data for said data structure of the data view, said method comprising the steps of:

(i) selecting a data component from a first existing data view;

(ii) indicating said selected data component to be added to a second data view, a target location of said selected data component in said second data view indicating a role of said selected data component in said second data view, said role being dependent on said visualisation type associated with said second data view;

(iii) creating a query for said second data view, said query specifying how to obtain data for said data structure for said second data view and being formed using said indicated data component and an interpretation of its corresponding role in said second data view; and

(iv) adding said selected data component to said second data view.

27. A method according to claim 26 wherein said created query represents a modification of an existing query for said second data view from the inclusion of said indicated data component.

28. A method according to claim 26 wherein said data structure is an hierarchical data structure and originates from at least one data source.

29. A method according to claim 28 wherein said data structure is represented using XML

30. A method according to claim 29 wherein said data originates from two or more heterogeneous data sources.

31. A method according to claim 26 wherein said existing query data are obtained using at least one of:

(a) electronic addresses of one or more data sources, said data sources being associated with stored data; and

(b) one or more definitions of existing views of data.

32. A method according to claim 26 further comprising the step of obtaining data for said generated query, said obtaining not requiring the presence of definitions of views of data that were used to generate said query, said generated query being defined in terms of data sources.

33. A method according to claim 26, further comprising a step, preceding step (i) of:

(i-a) establishing that a new data view is to be created, said new data view being assigned a visualisation type, wherein the visualisation type of said new data view is selected from a predetermined set of visualisation types.

34. A method according to claim 33 wherein said new data view is established by at least one of copying or moving said selected data component to a location in a workspace of said graphical user interface where said new data view is to be located, said visualisation type of said new data view being assigned that of said first data view from which said selected data component was selected.

35. A method according to claim 26 wherein said selected data component is one of copied to said second data view, or moved to said second data view.

36. A method according to claim 28 wherein said visualisation type is one of a table, a graph such as a line graph, a bar graph, or a pie graph, a scatter plot, a tree, and a grid display.

37. A method according to claim 28 wherein said selected data component is identified by a node address in said hierarchical data structure of said first data view.

38. A method according to claim 28 wherein said selected data component is identified by a node iterator and a node path with respect to said node iterator in said hierarchical data structure of said first data view, said node iterator representing a node pattern that is repeated in said hierarchical data structure.

39. A method according to claim 38 wherein said selected data component is represented as one of:

a column of a table in said first data view;

a set of labels used in a pie chart in said first data view;

a set of labels used as an x-axis of a graph in said first data view;

a set of values used in a pie chart in said first data view; and

a data series of a graph in said first data view.

40. A method according to claim 38 wherein said selected data component is associated with a second data component, said second data component specifying a set of values which have point wise correspondence with the data of the selected data component.

41. A method according to claim 26 further comprising, between steps (ii) and (iii), the step of:

(ii-a) checking that said selected data component can assume said indicated role in said second data view before permitting said selected data component to be incorporated in said second data view.

42. A method according to claim 36 wherein said indicated role in said second data view is one of:

(1) an identified column position in said table;

(2) a data set corresponding to the x axis of said graph;

(3) a data series corresponding to a selected y-axis of said graph;

(4) a data set corresponding to labels of said pie graph;

(5) a data series corresponding to values of said pie graph;

(6) node corresponding to an identified node position in said tree;

(7) a data set corresponding to an identified position in a grid unit of said grid display; and

(8) an identified data series position for said visualisation type.

43. A method according to claim 26 further comprising the step of:

(v) modifying said generated query of said second data view by applying a further operation on data presented in said second data view, said further operation being indicated using said graphical user interface.

44. A method according to claim 43 wherein said operation comprises at least one of:

a filter operation;

a sort operation;

a transform operation;

a rename operation;

a hiding operation for hiding a data component, said hiding operation resulting in data for said hidden data component remaining in said data structure of said second data view but not being presented in said graphical user interface; and

a combining operation that combines the data from two or more data components.

45. A method according to claim 44 wherein said operation is defined by a user of said graphical user interface demonstrating the required operation by manipulating examples of said data.

46. A method according to claim 26 wherein step (i) inlcudes a dragging operation upon said selected data component from said first data view in said graphical user interface, and step (ii) inlcudes a dropping operation of said selected data upon said second data view in said graphical user interface.

47. A computer readable medium, having a program recorded thereon, where the program is configured to make a computer execute a procedure to form a query from manipulations of existing query data in a graphical user interface, said existing query data comprising one or more views of data, each said data view being a visualisation of a data structure and associated with a visualisation type and a query, said query defining how to obtain data for said data structure of the data view, said program comprising:

.

(i) code means for selecting a data component from a first existing data view;

(ii) code means for indicating said selected data component to be added to a second data view, a target location of said selected data component in said second data view indicating a role of said selected data component in said second data view, said role being dependent on said visualisation type associated with said second data view;

(iii) code means for creating a query for said second data view, said query specifying how to obtain data for said data structure for said second data view and being formed using said indicated data component and an interpretation of its corresponding role in said second data view; and

(iv) code means for adding said selected data component to said second data view.

48. Apparatus for forming a query from manipulations of existing query data in a graphical user interface, said existing query data comprising one or more views of data, each said data view being a visualisation of a data structure and associated with a visualisation type and a query, said query defining how to obtain data for said data structure of the data view, said apparatus comprising:

means for selecting a data component from a first existing data view;

means for indicating said selected data component to be added to a second data view, a target location of said selected data component in said second data view indicating a role of said selected data component in said second data view, said role being dependent on said visualisation type associated with said second data view;

means for creating a query for said second data view, said query specifying how to obtain data for said data structure for said second data view and being formed using said indicated data component and an interpretation of its corresponding role in said second data view; and

means for adding said selected data component to said second data view.