System and method for calibrating and extrapolating management-inherent complexity metrics and human-perceived complexity metrics of information technology management

• US 8,001,068 B2
• Filed: 06/05/2006
• Issued: 08/16/2011
• Est. Priority Date: 06/05/2006
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Keller, Alexander, Diao, Yixin, Kearney, Robert D., Filepp, Robert
• Primary Examiner(s): VINCENT, DAVID ROBERT
• Application Number:
  US11/422,195
• Publication Number:
  US 20070282644A1
• Time in Patent Office: 1,898 Days
• Field of Search: 706/62, 706/12, 706 45- 48
• US Class Current: 706/45
• CPC Class Codes: G06Q 10/063 Operations research, analysis or management
• CPC Class Codes: G06Q 10/06395 Quality analysis or management
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

The invention broadly and generally provides a method for calibrating the relationship between management-inherent complexity metrics deriving from the management structure and human perceived complexity of information technology management comprising: (a) obtaining a set of management-inherent complexity metrics; (b) obtaining a set of human-perceived complexity metrics; (c) constructing a control model identifying a set of dominant indicators selected from the aforesaid set of management-inherent complexity metrics; (d) establishing a value model mapping from the aforesaid set of dominant indicators to the aforesaid set of human-perceived complexity metrics.

Claims

1. A method for calibrating the relationship between management-inherent complexity metrics deriving from the management structure of a system and human-perceived complexity of information technology management, said method comprising:

obtaining a set of management-inherent complexity metrics from quantified results of a complexity analysis, the complexity analysis quantifying a complexity of a configuration procedure between the system and an administrator of the system;

obtaining a set of human-perceived complexity metrics from user studies from the administrator of the system;

constructing a control model on a first processor, said control model identifying a set of dominant indicators selected from said set of management-inherent complexity metrics;

establishing a value model on a second processor, said value model mapping from said set of dominant indicators selected from said set of management-inherent complexity metrics to said set of human-perceived complexity metrics.

2. The method as set forth in claim 1, further comprising:

obtaining and validating said control model and said value model for quality assessment; and

repeating said obtaining and validating said control model and said value model for quality assessment.

3. The method as set forth in claim 1, wherein said set of management-inherent complexity metrics comprise at least one of:

execution complexity metrics;

parameter complexity metrics; and

memory complexity metrics.

4. The method as set forth in claim 1, wherein said value model is constructed using at least one of:

a statistical approach;

linear regression;

machine learning; and

an artificial neural network, wherein said artificial neural network is a radial basis function.

5. The method as set forth in claim 1, wherein said step of constructing a control model comprises at least one of:

(a) obtaining a subset of management-inherent complexity metrics as a set of dominant indicators under study;

(b) constructing a value model from said set of dominant indicators and said set of human-perceived complexity metrics based on a set of information technology management data; and

(c) evaluating the quality of said value model based on a different set of information technology management data.

6. The method as set forth in claim 5, further comprising obtaining a different subset of management-inherent complexity metrics from said set of dominant indicators under study.

7. The method as set forth in claim 6, further comprising repeating said step of obtaining a different subset of management-inherent complexity metrics as said set of dominant indicators under study until no better set of dominant indicators is found.

10. The method as set forth in claim 1, further comprising:

inputting the management-inherent complexity metrics into the value model; and

outputting a second set of human-perceived complexity metrics from the value model.

11. The method as set forth in claim 10, wherein at least one of the set of human-perceived complexity metrics and the second set of human-perceived complexity metrics each comprise at least one of cost metrics, efficiency metrics, and quality metrics.

12. The method as set forth in claim 10, further including inputting into the value model uncontrolled environment metrics to classify different information technology management types, the uncontrolled environment metrics comprising process-related metrics and process-independent metrics, the process-related metrics including at least one of goal metrics, prerequisite metrics, and scenario metrics, the process-independent metrics including at least one of preference metrics and fatigue metrics.

13. The method as set forth in claim 4, wherein inputs of said linear regression comprise a number of actions and a number of context switches, wherein outputs of said linear regression comprise an execution time, and wherein coefficients of said linear regression are obtained using a least squares approach.

14. The method as set forth in claim 4, wherein said radial basis function builds a nonlinear relationship, the radial basis function comprising environment variables to classify different information technology management types.

8. A method for extrapolating from management-inherent complexity metrics to human-perceived complexity of information technology management, said method comprising:

collecting a set of management-inherent complexity metrics from quantified results of a complexity analysis, the complexity analysis quantifying a complexity of a configuration procedure between the system and an administrator of the system;

obtaining a value model;

predicting human-perceived complexity with a processor, the human-perceived complexity being based on said set of management inherent complexity metrics and said value model, said predicting including;

inputting the management-inherent complexity metrics into the value model; and

outputting human-perceived complexity metrics from the value model.

15. The method as set forth in claim 8, wherein said management-inherent complexity metrics comprise at least one of:

execution complexity metrics;

parameter complexity metrics; and

memory complexity metrics.

16. The method as set forth in claim 8, wherein the human-perceived complexity metrics comprise at least one of cost metrics, efficiency metrics, and quality metrics.

17. The method as set forth in claim 8, further including inputting into the value model uncontrolled environment metrics to classify different information technology management types, the uncontrolled environment metrics comprising process-related metrics and process-independent metrics, the process-related metrics including at least one of goal metrics, prerequisite metrics, and scenario metrics, the process-independent metrics including at least one of preference metrics and fatigue metrics.

18. The method as set forth in claim 8, wherein said obtaining of the value model comprises constructing a linear regression model, wherein inputs of said linear regression model comprise a number of actions and a number of context switches, wherein outputs of said linear regression model comprise an execution time, and wherein coefficients of said linear regression model are obtained using a least squares approach.

19. The method as set forth in claim 8, wherein said obtaining of the value model comprises constructing a radial basis function neural network for building a nonlinear relationship, the radial basis function neural network comprising environment variables to classify different IT management types.

9. A program storage device readable by a digital processing apparatus and having a program of instructions which are tangibly embodied on the storage device and which are executable by the processing apparatus to perform a method for calibrating the relationship between management-inherent complexity metrics deriving from the management structure of a system and human-perceived complexity of information technology management, said method comprising:

obtaining a set of management-inherent complexity metrics from quantified results of a complexity analysis, the complexity analysis quantifying a complexity of a configuration procedure between the system and an administrator of the system;

obtaining a set of human-perceived complexity metrics from user studies from the administrator of the system;

constructing a control model on a first processor, said control model identifying a set of dominant indicators selected from said set of management-inherent complexity metrics; and

establishing a value model on a second processor, said value model mapping from said set of dominant indicators to said set of human-perceived complexity metrics.

20. The program storage device as set forth in claim 9, further comprising:

inputting the management-inherent complexity metrics into the value model, wherein the management-inherent complexity metrics comprises;

execution complexity metrics,

parameter complexity metrics, and

memory complexity metrics; and

outputting a second set of human-perceived complexity metrics from the value model, wherein the first set of human-perceived complexity metrics and the second set of human-perceived complexity metrics each comprises;

cost metrics,

efficiency metrics, and

quality metrics.