Conducting and managing sampled information audits for the determination of database accuracy

• US 5,404,509 A
• Filed: 05/08/1992
• Issued: 04/04/1995
• Est. Priority Date: 05/08/1992
• Status: Expired due to Fees

• Current Assignee: Renaissance Group IP Holdings, LLC
• Original Assignee: Laurence C. Klein
• Inventors: Klein, Laurence C.
• Primary Examiner(s): Black, Thomas G.
• Assistant Examiner(s): Lintz, Paul R.
• Application Number:
  US07/880,397
• Time in Patent Office: 1,061 Days
• Field of Search: 395/600, 395/575, 371/15.1, 364/401, 364/468, 364/408
• US Class Current: 1/1
• CPC Class Codes: G06F 11/008 Reliability or availability analysis
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

A computer-based method and apparatus for auditing electronic information, most often a database. A database auditor of the present invention conducts an audit as specified by a user-defined project. The project indicates focus groups, filters, skews and whether to count blank entries. The database auditor selects a sample representative of the view of the database described by the project. It presents the sample to the user in a standardized set of reports or on-line forms. The user then determines the number errors contained in the sample and communicates these data to the database auditor. The database auditor uses the error data to calculate the accuracy of the database, as well as the accuracies of the individual fields and focus groups, and to presents these accuracies to the user. Finally, the database auditor charts areas of accuracy and inaccuracy by field and focus group and indicates which inaccuracies are due to process errors and which are due to user errors. The indication of whether the source of inaccuracy is inherent in the process (i.e., a process error) or caused by human negligence (i.e., a user error) enables the user to efficiently and effectively correct database inaccuracies.

Claims

1. In a computer based system, an apparatus for auditing a repository of data stored electronically in the computer, wherein the repository of data stores a plurality of records each having a one or more fields, the apparatus comprising:

a conducting audit utility comprising;

audit criteria accessing means for accessing audit criteria stored in the computer, wherein said audit criteria includes the number of items stored in the repository of data to be audited;

sample generation means for generating a sample set of data based on said audit criteria, such that said sample set of data includes at least the number of items to audit as indicated in said audit criteria, wherein said sample set of data is selected by applying at least one of a focus group criteria, a filter criteria, a skew criteria, or an empty field indicator, wherein sample generation means includes,

focus group means, responsive to said focus group criteria, for logically organizing a variety of fields within the repository of data whose combined accuracy should be analyzed as one unit,

filter means, responsive to said filter criteria, for determining records and fields for inclusion in said sample set,

empty field means, responsive to said empty field indicator, for not including empty fields in said sample set when said sample set is generated, and

skew means, responsive to said skew criteria, for emphasizing one or more fields within a record such that sample set is biased towards said emphasized one or more fields, but does not limit said sample set to only emphasized fields,

a reviewing sample utility comprising means for providing said sample set of data in a user readable format so that a reviewer can compare said sample set of data with the original data stored in the repository of data and tabulate a total number of errors in said sample set of data; and

an error analysis utility comprising means for accepting said errors, and calculating audit results, whereby said audit results indicates accuracy values of said repository of data.

2. The apparatus of claim 1, wherein said sample generation means comprises:

means for identifying at least one field of the repository which is in said focus group of said audit;

means for accessing configuration data of said focus group, said configuration data comprising a target accuracy, a minimum accuracy, a confidence level, a sample size, a p0 value, a p1 value and an indication of whether to include said empty fields; and

means for accessing said filter criteria, said skew criteria and a skew emphasis factor for at least one field of said audit.

3. The apparatus of claim 1, wherein said sample generating means comprises:

generating means for generating a filtered list by applying said filter criteria associated with fields in the repository to select a portion of the repository;

selecting means for selecting a random record from said filtered list;

applying means for applying said skew criteria associated with fields in the repository to determine whether said random record should be included in said sample set; and

means for repeatedly invoking said generating means, said selecting means and said applying means until said sample set meets said audit criteria.

4. The apparatus of claim 1, wherein said reviewing sample utility comprises means for generating in a user readable format:

a predetermined number of fields selected for said audit,

a predetermined number of filled fields in each field in said sample set,

a predetermined number of filled fields in each group of fields in said sample set,

a predetermined number of empty fields in each field in said sample set, and

a predetermined number of empty fields in each group in said sample set.

5. The apparatus of claim 1, wherein said reviewing sample utility further comprises presenting means for presenting an on-line set of forms that enables said user to cycle through records in said sample set.

6. The apparatus of claim 5, wherein said presenting means comprises means for presenting images stored on the computer based system associated with said records in said sample set.

7. The apparatus of claim 1, wherein said reviewing sample utility further comprises receiving means for presenting reports and on-line forms that enable said user to input said feedback on the number of errors in said sample set.

8. The apparatus of claim 1, further comprising:

means for calculating an accuracy for each field of said audit;

means for calculating an accuracy of each focus group of said audit; and

means for calculating an overall accuracy of said audit.

9. The apparatus of claim 1, further comprising:

means for displaying field accuracies, and

means for denoting fields having inaccuracies due to process errors and fields having inaccuracies due to user errors.

10. The apparatus of claim 1, further comprising:

means for displaying focus group accuracies, and

means for denoting focus groups having inaccuracies due to process errors and focus groups having inaccuracies due to user errors.

11. An apparatus in a computer based system for auditing a repository of data stored electronically in the computer, comprising:

a project manager facility comprising an audit criteria utility comprising;

means for receiving audit criteria that includes at least one of a focus group criteria, a filter criteria, an empty field indicator, or a skew criteria; and

means for defining a project according to said audit criteria; and

an audit facility comprising;

a conducting audit utility comprising;

means for initiating an audit; and

sample generating means for generating a sample set of data comprising less than all of the stored data so as to be consistent with said audit criteria, wherein sample generating means includes,

focus group means, responsive to said focus group criteria, for logically organizing a variety of fields within repository of data whose combined accuracy should be analyzed as one unit,

filter means, responsive to said filter criteria, for determining records and fields field for inclusion said sample set,

empty field means, responsive to said empty field indicator, for not including empty fields in said sample set when said sample set is generated, and

skew means, responsive to said skew criteria, for emphasizing one or more fields within a record such that said sample set is biased towards said emphasized one or more fields, but does not limit said sample set to only emphasized field;

a reviewing sample utility comprising;

sample providing means for providing said sample set in a user readable format; and

means for receiving feedback on the number of errors in said sample set; and

an audit reporting and analysis utility comprising;

calculating means for calculating an audit result from said sample set based on said feedback; and

result providing means for providing said audit result in a user readable format.

12. The apparatus of claim 11, wherein said receiving means of said project manager facility comprises a project descriptive information utility which receives descriptive information about said project.

13. The apparatus of claim 11, wherein said receiving means of said project manager facility comprises an audit criteria utility which receives:

criteria to apply to a focus group within said project, and

at least one field of the repository to be included in said focus group.

14. The apparatus of claim 13, wherein said audit criteria utility comprises:

means for receiving a target accuracy, a minimum accuracy, and a confidence level for said a group; and

means for calculating a sample size, a reject level, a PO value and a P1 value for said focus group from said target accuracy, said minimum accuracy, and said confidence level.

15. The apparatus of claim 13, wherein said audit criteria utility comprises means for receiving additional criteria to apply to at least one additional a group within said project.

16. The apparatus of claim 13, wherein said audit criteria utility comprises:

means for receiving said filter criteria for said field; and

means for receiving at least one additional filter criteria for at least one additional field.

17. The apparatus of claim 13, wherein said audit criteria utility comprises:

means for receiving said skew criteria for said field; and

means for receiving at least one additional skew criteria for at least one additional field.

18. The apparatus of claim 17, further comprising means for receiving an emphasis value indicative of an extent to which said sample set is biased toward values of said field which meet said skew criteria.

19. The apparatus of claim 11, wherein said project manager facility comprises an audit manager utility for managing a set of audit results associated with said project, comprising:

means for adding an additional audit result to said set in response to an insert command; and

means for removing an existing audit result from said set in response to a delete command.

20. The apparatus of claim 19, wherein said audit manager utility further comprises means for automatically adding said audit result to said set if said audit is initiated from said project.

21. The apparatus of claim 19, wherein said audit manager utility further comprises:

means for recalculating an average accuracy, a median accuracy and an absolute accuracy of said set of audits upon adding said additional audit result to said set; and

means for recalculating said average accuracy, said median accuracy and said absolute accuracy of said set of audits upon removing said existing audit result from said set.

22. The apparatus of claim 11, wherein said initiating means of said conducting audit facility comprises means for initiating an audit according to said audit criteria defined as part of said project.

23. The apparatus of claim 11, wherein said initiating means of said conducting audit facility comprises means for initiating an audit according to current audit criteria not defined as part of said project.

24. The apparatus of claim 11, wherein said sample generating means comprises:

generating means for generating a filtered list by applying said filter criteria associated with fields in the repository to select a portion of the repository;

selecting means for selecting a random record from said filtered list;

applying means for applying said skew criteria associated with fields in the repository to determine whether said random record should be included in said sample set; and

means for repeatedly invoking said selecting means and said applying means until said sample set meets said audit criteria.

25. The apparatus of claim 11, wherein said conducting audit utility further comprises means for generating in a user readable format:

a predetermined number of fields selected for said audit,

a predetermined number of filled fields in each field in said sample set,

a predetermined number of filled fields in each focus group of fields in said sample set,

a predetermined number of empty fields in each field in said sample set, and

a predetermined number of empty fields in each focus group in said sample set.

26. The apparatus of claim 11, wherein said presenting means of said reviewing sample utility comprises means for presenting an on-line set of forms that enables said user to cycle through records in said sample set.

27. The apparatus of claim 11, wherein said presenting means of said reviewing sample utility further comprises means for presenting images stored on the computer based system associated with said records in said sample set.

28. The apparatus of claim 11, wherein said result providing means comprises means for facilitating entry of said feedback by said user by presenting reports and on-line forms that enable said user to input said feedback.

29. The apparatus of claim 11, wherein said calculating means comprises:

means for calculating an accuracy of each field of said audit;

means for calculating an accuracy of each focus group of said audit; and

means for calculating an overall accuracy of a portion of the repository specified by said audit criteria.

30. The apparatus of claim 11, wherein said result providing means comprises means for presenting an analysis report that

displays field accuracies, and

denotes fields having inaccuracies due to process errors and fields having inaccuracies due to user errors.

31. The apparatus of claim 11, wherein said result providing means comprises means for presenting an analysis report that

displays group accuracies, and

denotes focus groups having inaccuracies due to process errors and focus groups having inaccuracies due to user errors.

32. In a computer based system having an audit facility utility, a method for auditing a repository of data stored electronically in the computer, comprising the steps of:

(a) selecting in the audit facility utility a sample set of data which comprises less than all of the stored data and which is consistent with audit criteria stored in the computer based system; and

wherein said sample set of data is selected by applying at least one of a focus group criteria, a filter criteria, a skew criteria, or an empty field indicator, which includes the steps of;

(i) logically organizing, in responses to said focus group criteria, a variety of fields within the repository of data whose combined accuracy should be analyzed as one unit,

(ii) determining, in response to said filter criteria, records and fields for inclusion in said sample set,

(iii) not including, in response to said empty field indicator, empty fields in said sample set when said sample set is generated, and

(iv) emphasizing, in response to said skew criteria, one or more fields within a record such that said sample set is biased towards said emphasized one or more fields, but does not limit said sample set to only emphasized fields; and

(b) providing said sample set of data in a user readable format.

33. The method of claim 32, wherein step (a) comprises the steps of:

(c) applying said filter criteria to the repository at run time to generate a filtered list;

(d) randomly generating a number that identifies exactly one record in said filtered list which has not yet been added to said sample set;

(e) determining whether said record meets a skew requirement;

(f) branching to step (d) if said record does not meet said skew requirements;

(g) adding said record to said sample set;

(h) determining whether said sample set meets audit criteria requirements; and

(i) branching to step (c) if said sample set does not meet said audit criteria requirements.

34. The method of claim 32, wherein step (a) comprises the steps of:

(c) determining that a field in a record meets said skew requirement if;

(i) a value of said field meets said skew criteria for said field, or

(ii) a number of consecutive attempts to meet said skew criteria for said field is greater than an emphasis value of said field;

(d) determining that said record meets said skew requirement if each field in said record meets said skew requirement; and

(e) adding said record to said sample set if said record meets said skew requirement.

35. The method of claim 32, wherein step (a) comprises the steps of:

(c) selecting a field of said record;

(d) incrementing an empty field count associated with said selected field if an instance of said selected field in said record is empty;

(e) incrementing a filled field count associated with said selected field if said instance of said selected field in said record is filled;

(f) branching to step (c) if there is an unselected field in said record;

(g) adding to a group empty count associated with a focus group said empty field count associated with each said field in said focus group, wherein said group specifies a set of fields in the repository;

(h) adding to a group filled count associated with said focus group said filled field count associated with each said field in said focus group;

(i) branching to step (g) if said group empty count and said each group filled count has not been calculated for any focus group which has been defined; and

(j) using each group empty count and group filled count to determine whether said audit criteria are met for each focus group.

36. The method of claim 35, wherein step (j) comprises the steps of:

(k) selecting one said focus group;

(l) setting a current sample count for said selected focus group to said filled field count of said selected focus group;

(m) adding to said current sample set of said selected focus group said empty field count of said selected focus group if empty fields are counted for said selected focus group;

(n) determining that said sample set meets said audit criteria for said selected focus group if a minimum sample set size of said selected focus group is less than said current sample count of said selected focus group; and

(o) branching to step (k) if any focus group which has been defined has not been selected in step (k).

37. The method claim 32, wherein step (b) comprises the step of presenting a standardized set of templates of said sample set to enable a user to compare the information in said sample set to an original source of the information.

38. The method of claim 37, further comprising the step of presenting said original image of said original source if said image is stored in the computer.

39. In a computer based system having a project manager utility and an audit facility utility, a method for auditing a repository of data stored electronically in the computer, comprising the steps of:

(a) inputting audit criteria to the project manager facility utility;

(b) selecting in the audit facility utility a sample set of data which comprises less than all of the stored data and which is consistent with said audit criteria, wherein said sample set of data is selected by applying at least one of a focus group criteria, a filter criteria, a skew criteria, or an empty field indicator, which includes the steps of;

(i) logically organizing, in response to said focus group criteria, a variety or fields within the repository of data whose combined accuracy should be analyzed as one unit,

(ii) determining, in response to said filter criteria, records and fields for inclusion in said sample set,

(iii) including, in response to said empty field indicator, empty fields in said sample set when said sample set is generated, and

(iv) emphasizing, in response to said skew criteria, one or more fields within a record such that said sample set is biased towards said emphasized one or more fields, but does not limit said sample set to only emphasized fields;

(c) providing said sample set of data in a user readable format;

(d) inputting error information to the audit facility utility based on a user review of the sample set of data;

(e) determining in the audit facility utility an accuracy rate of the sample set of data and extrapolating said accuracy rate to the entire repository of data; and

(f) providing a result determined in step (e) in a user readable format.

40. The method claim 39, further comprising the steps of repeating steps (a)-(f).

41. The method of claim 39, further comprising the steps of repeating steps (b)-(f) so as to audit the repository according to previously defined audit criteria.

42. The method of claim 39, wherein step (a) comprises the steps of:

(g) inputting to the project manager facility group accuracy requirements applied to a focus group, wherein said focus group comprises at least one field of the repository; and

(h) determining audit statistical requirements from said group accuracy requirements.

43. The method of claim 42, further comprising the step of repeating (g) and (h).

44. The method of claim 42, wherein step (g) comprises the step of inputting to the project manager facility a target accuracy value, a minimum accuracy value, and a confidence level value so as to enable said audit statistical requirements for said focus group to be determined in step (h) without consulting the repository.

45. The method of claim 39, wherein step (a) comprises the steps of:

(g) inputting into the project manager facility said filter criteria associated with a field in the repository; and

(h) repeating step (g) if there is an additional filter criteria associated with an additional field in the repository.

46. The method of claim 39, wherein step (a) comprises the steps of:

(g) inputting into the project manager facility said skew criteria associated with a field in the repository; and

(h) repeating step (g) if there is an additional skew criteria associated with an additional field in the repository.

47. The method of claim 46, wherein step (g) further comprises the step of inputting into the project manager facility an emphasis value indicative of an extent to which said sample set is biased toward values of said field which meet said skew criteria.

48. The method of claim 39, wherein step (b) comprises the steps of:

(g) applying said filter criteria to the repository at run time to generate a filtered list, wherein said filter criteria is in said audit criteria;

(h) randomly generating a number that identifies exactly one record in said filtered list which has not yet been added to said sample set;

(i) determining whether said record meets a skew requirement, wherein said skew requirement is in said audit criteria;

(j) branching to step (h) if said record does not meet said skew requirement;

(k) adding said record to said sample set;

(l) determining whether said sample set meets requirements of said audit criteria; and

(m) branching to step (g) if said sample set does not meet said requirements of said audit criteria.

49. The method of claim 39, wherein step (b) comprises the steps of;

(g) selecting a record of the repository;

(h) determining that an instance of a field in said record meets a skew requirement if;

said instance of said field meets said skew criteria of said field, wherein said skew criteria is in said audit criteria, or

(ii) instances of said field in a number of consecutive records previously selected failed to meet said skew criteria, wherein said number is greater than an emphasis value of said field, wherein said emphasis count is in said audit criteria;

(i) determining that said selected record meets said skew requirement if each field in said selected record meets said skew requirement; and

(j) adding said record to said sample set if said record meets said skew requirements.

50. The method of claim 39, wherein step (b) comprises the step of:

(g) adding to said sample a record which meets said audit criteria;

(h) selecting a field of said record;

(i) incrementing an empty field count associated with said field if an instance of said field is empty;

(j) incrementing a filled field count associated with said field if said instance is not empty;

(k) branching to step (g) if there is an unselected field in said record;

(l) adding said empty field count associated with each said field in a focus group to a group empty count associated with said focus group, wherein said focus group specifies a set of at least one field in the repository;

(m) adding said filled field count associated within each said field in said focus group to a group filled count associated with said a group;

(n) branching to step (l) if said group empty count and said group filled count has not been calculated for any focus group which has been defined; and

(o) using each group empty count and each group filled count to determine whether said audit criteria are met for each focus group.

51. The method of claim 50, wherein step (o) comprises the steps of:

(p) selecting one said focus group;

(q) selecting a current sample count for said selected focus group to said filled field count of said selected focus group;

(r) adding to said current sample set of said selected focus group said empty field count of said selected focus group if empty fields are counted for said selected focus group;

(s) determining that said sample set meets said audit criteria for said selected focus group if a minimum sample set size of said selected focus group is less than said current sample count of said selected focus group; and

(t) branching to step (p) if any focus group which has been defined has not been selected in step (p).

52. The method of claim 39, wherein step (c) comprises the step of presenting a standardized set of templates of said sample set to enable a user to compare the information in said sample set to an original source of the information.

53. The method of claim 52, further comprising the step of presenting said original image of said original source if said image is stored in the computer.

54. The method of claim 39, wherein step (d) comprises the step of presenting a standardized set of templates of said sample set, so as to enable a user to communicate discrepancies between said sample set and said original source.

55. The method of claim 39, wherein step (e) comprises the steps of:

(g) calculating a field level accuracy indicative of the accuracy of each field in the repository;

(h) calculating a group level accuracy indicative of the accuracy of each group of fields; and

(i) calculating a repository accuracy indicative of the accuracy of the repository.

56. The method of claim 55, wherein step (g) comprises the steps of:

(j) selecting one said field;

(k) setting a current field sample count for said selected field to a quantity of filled fields for said selected field; and

(l) adding to said current field sample count a quantity of empty fields for said selected field if empty fields are to be counted for said focus group to which said selected field is assigned;

(m) dividing said current field sample count by a quantity of errors in said field, wherein said quantity of errors is determined from said error information of step (d); and

(n) repeating steps (j)-(m) until each field in the repository has been selected.

57. The method of claim 55, wherein step (h) comprises the steps of:

(j) selecting one said focus group;

(k) setting a current group sample count for said selected focus group to a quantity of filled fields in said selected focus group;

(l) adding to said current group sample count a quantity of empty fields in said selected focus group if empty fields are to be counted for said selected focus group;

(m) dividing said current group sample count by a quantity of errors in said selected focus group, wherein said quantity of errors is determined from said error information of step (d); and

(n) repeating steps (j)-(m) until each fosu group in the repository has been selected.

58. The method of claim 55, wherein step (i) comprises the steps of:

(j) determining a quantity of fields in the repository;

(k) determining from said error information of step (d) a quantity of said fields which are erroneous; and

(l) dividing said quantity of fields in the repository by said quantity of fields which are erroneous.

59. The method of claim 39, wherein step (f) comprises the step of generating a standardized set of templates of said sample set so as to present said audit result to said user in detail and in summary form.

60. The method of claim 39, further comprising the step:

(g) of computing accuracy results of a plurality of audits.

61. The method of claim 60, further comprising the steps of:

(h) computing an average accuracy of said audit;

(i) computing an median accuracy of said audit; and

(j) computing an absolute accuracy of said audit.

62. The method of claim 39, wherein step (f) comprises the steps of:

(g) presenting a plot of accuracies of fields of the repository; and

(h) presenting a demarcation range indicative of a process accuracy rate of said fields, so as to distinguish between process errors and user errors for each of said fields.

63. The method of claim 39, wherein step (f) comprises the steps of:

(g) presenting a plot of accuracies of focus groups of the repository; and

(h) presenting a demarcation range indicative of a process accuracy rate of said focus groups, so as to distinguish between process errors and user errors for each of said focus groups.