Smarter business intelligence systems

• US 8,688,606 B2
• Filed: 01/24/2011
• Issued: 04/01/2014
• Est. Priority Date: 01/24/2011
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Negi, Sumit, Bhide, Manish Anand, Batra, Vishal Singh, Kothari, Govind
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): FINK, THOMAS ANDREW
• Application Number:
  US13/012,554
• Publication Number:
  US 20120191636A1
• Time in Patent Office: 1,163 Days
• Field of Search: 706/46, 709/223, 709/224
• US Class Current: 706/45
• CPC Class Codes: G06Q 10/00 Administration; Management

Abstract

An embodiment of the invention provides a method and system for analyzing a plurality of reports. More specifically, a change detection module predicts results of future reports based on past reports and identifies a first report that deviates from its predicted results. A dependency analysis module connected to the change detection module at least one report sharing a dependency with the first report by performing a dependency analysis and/or a usage analysis. The dependency analysis labels the first report and at least one second report as sharing a dependency if the second report deviates from its predicted results. The usage analysis labeling the first report and at least one report analyzed by an analyst as sharing a dependency if the report analyzed by the analyst is analyzed in response to the identification of the first report.

Claims

1. A method executed by a processor for analyzing a plurality of reports, said method comprising:

predicting results of future reports based on past reports;

identifying a first report that deviates from its predicted results with a change detection module; and

performing a dependency analysis and a usage analysis with a dependency analysis module,

said performing of the dependency analysis including;

identifying at least one second report that deviates from its predicted results, and

labeling the first report and the at least one second report as sharing a dependency, and

said performing of the usage analysis including;

identifying at least one report that has previously been analyzed by a human analyst, said at least one report that has previously been analyzed by a human analyst having been requested by the human analyst in response to said identifying of the first report, and

subsequent to said identifying at least one report that has previously been analyzed by a human analyst, labeling the first report and the at least one report analyzed by the human analyst as sharing a dependency based on the fact that said at least one report that has previously been analyzed by a human analyst was requested by the human analyst in response to said identifying of the first report.

2. The method according to claim 1, further comprising:

when the dependency analysis is performed, identifying at least one sub-component of the first report and at least one sub-component of the at least one second report that share a dependency; and

when the usage analysis is performed, identifying at least one sub-component of the first report and at least one sub-component of the at least one report analyzed by the human analyst that share a dependency.

3. The method according to claim 1, further comprising performing a root cause analysis for a select report, the root cause analysis including:

analyzing at least one dependent report, the dependent report sharing a dependency with the select report, said analyzing of the at least one dependent report including calculating a deviation for each of the sub-components of the dependent report that deviate from the predicted results of the dependent report; and

ranking the sub-components of the dependent report based on the calculated deviations.

4. The method according to claim 3, wherein the select report is the first report, the at least one second report, or the at least one report analyzed by the human analyst.

5. The method according to claim 3, further comprising, when the at least one dependent report has a threshold deviation from the predicted results of the dependent report:

analyzing at least one second dependent report, the second dependent report sharing a dependency with the dependent report, said analyzing of the at least one second dependent report including calculating a deviation for each of the sub-components of the second dependent report that deviate from the predicted results of the second dependent report; and

ranking the sub-components of the dependent report and the sub-components of the second dependent report based on the calculated deviations.

6. A method executed by a processor for analyzing a plurality of reports, said method comprising:

predicting results of future reports based on past reports;

identifying a first report that deviates from its predicted results with a change detection module;

performing a dependency analysis and a usage analysis with a dependency analysis module,

said performing of the dependency analysis including, labeling the first report and at least one second report as sharing a dependency if the at least one second report deviates from its predicted results, and

said performing of the usage analysis including identifying at least one report that has previously been analyzed by a human analyst, said at least one report that has previously been analyzed by a human analyst having been requested by the human analyst in response to said identifying of the first report, and subsequent to said identifying at least one report that has previously been analyzed by a human analyst, labeling the first report and the at least one report analyzed by the human analyst as sharing a dependency based on the fact that said at least one report that has previously been analyzed by a human analyst was requested by the human analyst in response to said identifying of the first report; and

performing a fine grain analysis, including;

if the dependency analysis is performed, labeling the first report and the at least one second report as sharing a dependency if at least one sub-component of the first report and at least one sub-component of the at least one second report share a dependency, and

if the usage analysis is performed, labeling the first report and the at least one report analyzed by the human analyst as sharing a dependency if at least one sub-component of the first report and at least one sub-component of the at least one report analyzed by the human analyst share a dependency.

7. The method according to claim 6, further comprising performing a root cause analysis for a select report, the root cause analysis including:

analyzing at least one dependent report, the dependent report sharing a dependency with the select report, said analyzing of the at least one dependent report including calculating a deviation for each of the sub-components of the dependent report that deviate from the predicted results of the dependent report; and

ranking the sub-components of the dependent report based on the calculated deviations.

8. The method according to claim 7, wherein the select report is the first report, the at least one second report, or the at least one report analyzed by the human analyst.

9. The method according to claim 7, further comprising, when the at least one dependent report has a threshold deviation from the predicted results of the dependent report:

analyzing at least one second dependent report, the second dependent report sharing a dependency with the dependent report and lacking a dependency from the select report, said analyzing of the at least one second dependent report including calculating a deviation for each of the sub-components of the second dependent report that deviate from the predicted results of the second dependent report; and

ranking the sub-components of the dependent report and the sub-components of the second dependent report based on the calculated deviations.

10. A method executed by a processor for analyzing a plurality of reports, said method comprising:

predicting results of future reports based on past reports;

identifying a first report that deviates from its predicted results with a change detection module;

performing a dependency analysis and a usage analysis with a dependency analysis module,

said performing of the dependency analysis including;

identifying at least one second report that deviates from its predicted results, and

labeling the first report and the at least one second report as sharing a dependency, and

said performing of the usage analysis including;

identifying at least one report that has previously been analyzed by a human analyst, said at least one report that has previously been analyzed by a human analyst having been requested by the human analyst in response to said identifying of the first report, and

subsequent to said identifying at least one report that has previously been analyzed by a human analyst, labeling the first report and the at least one report that has previously been analyzed by the human analyst as sharing a dependency based on the fact that said at least one report that has previously been analyzed by a human analyst was requested by the human analyst in response to said identifying of the first report; and

performing a root cause analysis for a select report, the root cause analysis including;

analyzing at least one dependent report, the dependent report sharing a dependency with the select report, said analyzing of the at least one dependent report including calculating a deviation for each of the sub-components of the dependent report that deviate from the predicted results of the dependent report; and

ranking the sub-components of the dependent report based on the calculated deviations.