Drill-through queries from data mining model content

• US 20050021482A1
• Filed: 06/30/2003
• Published: 01/27/2005
• Est. Priority Date: 06/30/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: MacLennan, C. James, Tang, Zhaohui, Kim, Pyungchul, Iyer, Raman
• Granted Patent: US 7,188,090 B2
• Time in Patent Office: Days
• US Class Current: 706/11
• CPC Class Codes: G06F 16/2425 Iterative querying; Query formulation based on the results of a preceding query
• CPC Class Codes: G06F 16/2428 Query predicate definition using graphical user interfaces, including menus and forms G06F16/2423 takes precedence
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases

Abstract

A drill-through feature is provided which provides a universal drill-through to mining model source data from a trained mining model. In order for a user or application to obtain model content information on a given node of a model, a universal function is provided whereby the user specifies the node for a model and data set, and the cases underlying that node for that model and data set are returned. A sampling of underlying cases may be provided, where only a sampling of the cases represented in the node is requested.

Claims

1. A method for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said method comprising:

receiving a request for data underlying a specified node from among said at least one node;

determining which of said at least one case corresponds to said specified node; and

providing data regarding said corresponding cases.

2. The method of claim 1, where said receiving includes detecting a request for data underlying a specified node, said detecting comprising:

receiving a query having query information associated therewith; and

detecting that said query information comprises said request for data underlying a specified node.

3. The method of claim 2, where said receiving a query comprises receiving a query in a specific query language.

4. The method of claim 3, wherein said receiving a query includes receiving a query in SQL database query language.

5. The method of claim 1, where said providing data regarding said corresponding cases comprises providing data regarding each of said corresponding cases.

6. The method of claim 1, where said receiving a request comprises receiving a request for a sample of data, and where said providing data regarding said corresponding cases comprises providing data regarding a sampling of said corresponding cases.

7. A computer readable medium comprising computer executable modules having computer executable instructions for performing the method of claim 1.

8. A computer readable medium comprising computer executable modules having computer executable instructions, said modules for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said computer executable modules comprising:

an input component for receiving an indication of a node from among said at least one node;

an analysis component for determining which of said at least one case corresponds to said specified node; and

an output component that outputs data regarding said corresponding cases.

9. The computer readable medium of claim 8, where said input component detects a request for data underlying a specified node.

10. The computer readable medium of claim 9, where said input component receives a query and detects that said query includes said request for data underlying a specified node.

11. The computer readable medium of claim 10, where said input component receives a query a query in a specific query language.

12. The computer readable medium of claim 11, wherein said input component receives a query in SQL database query language.

13. The computer readable medium of claim 8, where said output component provides data regarding each of said corresponding cases.

14. The computer readable medium of claim 8, where said input component receives a request for a sample of data, and said output component outputs data regarding a sampling of said corresponding cases.

15. An application programming interface for use in connection with providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, wherein said application programming interface receives as input a request for data underlying a specified node from among said at least one node, sends a query to a database comprising said data to retrieve which of said at least one case corresponds to said specified node and outputs data regarding said corresponding cases.

16. The application programming interface of claim 15, where said application programming interface detects a request for data underlying a specified node by receiving a query and detecting that said query comprises said request for data underlying a specified node.

17. The application programming interface of claim 16, where said application programming interface sends a query to the database in a specific query language.

18. The application programming interface of claim 17, wherein said application programming interface sends a query in SQL database query language.

19. The application programming interface of claim 15, where said application programming interface outputs data regarding each of said corresponding cases.

20. The application programming interface of claim 15, wherein when said application programming interface receives a request for a sample of data, said application programming interface outputs data regarding a sampling of said corresponding cases.

21. The application programming interface of claim 20, wherein said application programming interface one of (a) formats the query the application programming interface sends to the database to retrieve a sampling and (b) samples full results of a standard query for all of the corresponding cases.

22. The application programming interface of claim 15, wherein said query is sent and said results are retrieved via at least one network.

23. A user interface exposing a universal function enabling a user to specify a node, via the user interface, from a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case and in response to said user specifying the node, the cases underlying that node for that model and data set are retrieved and displayed to the user.

24. The user interface of claim 23, where said user interface detects a request for data underlying the specified node, and in response sends a request for data underlying a specified node via an application programming interface.

25. The user interface of claim 24, where said application programming interface sends a query to the database in a specific query language.

26. The user interface of claim 25, wherein said application programming interface sends a query in SQL database query language.

27. The user interface of claim 23, where said user interface displays data regarding each of said corresponding cases.

28. The user interface of claim 24, wherein when said user interface receives a request for a sample of data relating to the specified node, said user interface sends to the application programming interface a request for data regarding a sampling of said corresponding cases.

29. The user interface of claim 28, wherein said user interface displays a sampling of the corresponding cases.

30. A system for providing information regarding a graphical representation of a mining model applied to a data set comprising at least one case, said system comprising:

an application programming interface, said application programming interface (a) receiving as input a request for data underlying a specified node from a graphical representation of a mining model applied to said data set, said graphical representation comprising at least one node corresponding to at least one of said at least one case, (b) sending a query comprising said data to retrieve, from among said at least one case, one or more corresponding cases which correspond to said specified node, and (c) responding to said request with data regarding said corresponding cases; and

a database for storing said data set, operably connected with said application programming interface, and for receiving said query and returning said data regarding said corresponding cases to said application programming interface.

31. The system of claim 30, where said query comprises a query in a specific query language.

32. The system of claim 31, wherein said receiving a query includes receiving a query in SQL database query language.

33. The system of claim 30, where said providing data regarding said corresponding cases comprises providing data regarding each of said corresponding cases.

34. The system of claim 30, where said request comprises a request for a sample of data, and where said data regarding said corresponding cases comprises data regarding a sampling of said corresponding cases.

35. A system for providing information regarding a graphical representation of a mining model applied to a data set, said system comprising:

a user interface, said user interface (a) receiving as input a request for data underlying a specified node from a graphical representation of a mining model applied to a data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, (b) sending a query comprising said data to retrieve, from among said at least one case, one or more corresponding cases which correspond to said specified node, and (c) responding to said request with data regarding said corresponding cases; and

a database for storing said data set, operably connected with said user interface, and for receiving said query and returning said data regarding said corresponding cases to said application programming interface.

36. The system of claim 35, where said query comprises a query in a specific query language.

37. The system of claim 36, wherein said receiving a query includes receiving a query in SQL database query language.

38. The system of claim 35, where said providing data regarding said corresponding cases comprises providing data regarding each of said corresponding cases.

39. The system of claim 35, where said request comprises a request for a sample of data, and where said data regarding said corresponding cases comprises data regarding a sampling of said corresponding cases.

40. The system of claim 35, where said request for data underlying a specified node from a graphical representation of a mining model comprises an indication that a user has selected said node.

41. The system of claim 40, where said indication comprises a mouse click.

42. A system for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said system comprising:

request receipt means for receiving a request for data underlying a specified node from among said at least one node;

correspondence determination means for determining which of said at least one case corresponds to said specified node; and

data provision means for providing data regarding said corresponding cases.

43. A method for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said method comprising:

receiving a request for data underlying a specified node from among said at least one node;

determining at least one hypothetical case corresponding to said specified node; and

providing data regarding said corresponding hypothetical cases.

44. The method of claim 43, where said receiving includes detecting a request for data underlying a specified node, said detecting comprising:

receiving a query having query information associated therewith; and

detecting that said query information comprises said request for hypothetical data corresponding to a specified node.

45. The method of claim 44, where said receiving a query comprises receiving a query in a specific query language.

46. The method of claim 44, wherein said receiving a query includes receiving a query in SQL database query language.

47. A computer readable medium comprising computer executable modules having computer executable instructions for performing the method of claim 43.

48. A computer readable medium comprising computer executable modules having computer executable instructions, said modules for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said computer executable modules comprising:

an input component for receiving an indication of a node from among said at least one node;

an analysis component for determining at least one hypothetical case corresponding to said specified node; and

an output component that outputs data regarding said corresponding hypothetical cases.

49. The computer readable medium of claim 48, where said input component detects a request for hypothetical data underlying a specified node.

50. The computer readable medium of claim 49, where said input component receives a query and detects that said query includes said request for hypothetical data underlying a specified node.

51. A system for providing information regarding a graphical representation of a mining model applied to a data set, said data set comprising at least one case, said graphical representation comprising at least one node corresponding to at least one of said at least one case, said system comprising:

request receipt means for receiving a request for hypothetical data corresponding to a specified node from among said at least one node;

hypothetical case provision means for providing at least one hypothetical case corresponding to said specified node; and

data provision means for providing data regarding said corresponding hypothetical cases.