Computer assisted benchmarking system and method using induction based artificial intelligence

• US 20030050814A1
• Filed: 03/07/2002
• Published: 03/13/2003
• Est. Priority Date: 03/08/2001
• Status: Active Grant

• Current Assignee: Deloitte Development LLC
• Original Assignee: Deloitte Development LLC
• Inventors: Curet, Olivier L., Stoneking, Michael D.
• Granted Patent: US 8,788,452 B2
• Time in Patent Office: Days
• US Class Current: 705/7
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/06393 Score-carding, benchmarking or key performance indicator [KPI] analysis
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

A system and method are provided for the collection of business performance data and the identification of patterns or rules from such data that are key predictors of future business performance. The performance data are preferably collected using one or several questionnaires containing a plurality of questions that probe into specific performance areas of companies. The questionnaires are used to collect responses applicable to a plurality of companies and the responses applicable to each company are stored in a database as separate company profiles to define the knowledge-base from which a rule induction engine may identify the key discriminators of business performance.

Claims

1. A method of benchmarking to rank and position the performance of an entity in specific areas by identifying key discriminators that drive toward specific levels of business performance comprising the steps of obtaining business performance information for a plurality of entities, the business performance information including a first set of performance criteria, each performance criterion of the first set of performance criteria having an associated value for scoring the performance of the plurality of entities, and applying rule induction techniques to evaluate the first set of performance criteria and associated values to identify an optimal subset of performance criteria that are the key discriminators.

2. The method of claim 1 further comprising the steps of obtaining the first set of business criteria and the associated values via a questionnaire containing a plurality of questions that probe into specific performance areas of the entities.

3. The method of claim 1 further comprising the step of calculating for each entity an outcome score which is based on an average score of all the values for that entity and ranking the plurality of entities based on the outcome score of each entity.

4. The method of claim 3 wherein a quartile ranking is used such that the highest 25 percent of outcome scores are ranked in the top quartile, the middle 50 percent of outcome scores are ranked in the inter-quartile and the lowest 25 percent of outcome scores are ranked in the bottom quartile.

5. The method of claim 1 further comprising the step of representing the optimal subset of performance criteria that are the key discriminators as a classification tree wherein a first performance criterion of the subset of performance criteria is identified as a root node and all other performance criterion from the subset of performance criteria are identified as sub-nodes.

6. The method of claim 5 wherein the rule induction technique applied to the first set of performance criteria and associated values comprises a CART algorithm.

7. The method of claim 1 wherein the first set of performance criteria and the associated numerical values for scoring the performance of the plurality of entities are stored in a database as separate company profiles to define the knowledge-base from which a rule induction engine may identify the key discriminators of business performance.

8. The method of claim 1 wherein the associated values comprise numerical values based on a scale between 1 and 5.

9. The method of claim 8 further comprising the step of calculating for each entity an outcome score which is based on an average score of all the numerical values for that entity and ranking the plurality of entities based on the outcome score of each entity.

10. A method for benchmarking the performance of business entities comprising the steps of obtaining from a plurality of business entities business performance information, the information including a data set of business performance questions and numerical responses, calculating for each business entity an outcome score which is based on an average of the numerical responses of the business entity, and applying a rule induction technique to the business performance information to identify an optimal subset of questions and numerical responses that are key indicators of the outcome scores of the plurality of business entities.

11. The method of claim 10 further comprising the steps of obtaining the data set of business performance questions and numerical responses via a questionnaire having a plurality of questions that probe into specific performance areas of the business entities.

12. The method of claim 10 further comprising the step of ranking the plurality of business entities based on the outcome score of each business entity.

13. The method of claim 10 wherein a quartile ranking is used such that the highest 25 percent of outcome scores are ranked in the top quartile, the middle 50 percent of outcome scores are ranked in the inter-quartile and the lowest 25 percent of outcome scores are ranked in the bottom quartile.

14. The method of claim 10 further comprising the step of representing the optimal subset of questions and numerical responses as a classification tree wherein a first question and numerical response from the subset of questions and numerical responses is identified as a root node and all other questions and numerical responses from the subset of questions and numerical responses are identified as sub-nodes.

15. The method of claim 14 wherein the rule induction technique applied to the business information comprises a CART algorithm applied to the entire data set of business performance questions and numerical responses.

16. The method of claim 10 wherein the data set of business performance questions and numerical responses are stored in a database as separate company profiles to define the knowledge-base from which a rule induction engine may identify the key discriminators of business performance.

17. A method for benchmarking the performance of business entities comprising the steps of obtaining from each of a plurality of business entities a data set of responsive scores to business performance questions, the responsive scores for each business entity representing a case, calculating for each case an average of all the responsive scores, ranking the average score for each case based a quartile ranking wherein the highest 25 percent of the average scores are ranked in the top quartile, the middle 50 percent of the average scores are ranked in the inter-quartile and the lowest 25 percent of the average scores are ranked in the bottom quartile, and applying a CART algorithm to the data set to identify a subset of questions and responses that are key indicators of the average scores of the plurality of business entities.

18. The method of claim 17 wherein the subset of questions define a tree structure having an index node and sub-nodes.

19. The method of claim 17 wherein the CART algorithm loops through each of the questions and the responsive scores to define the subset of questions.

20. The method of claim 17 further comprising the step of providing the plurality of business entities with a questionnaire including a plurality of questions and responsive scores between 1 and 5 are applied by the plurality of business entities.

21. The method of claim 17 further comprising the step of populating a database with the questions and responsive scores.

22. In a data set including business performance data for a plurality of business entities, wherein the business performance data includes responses to a plurality of business performance questions and the responses have a numerical value, a method of evaluating the data set to identify key questions that drive toward specific levels of business performance comprising the steps of applying the questions and responses to a rule induction engine, and cycling through the questions and responses with the rule induction engine to identify a subset of questions and responses that identify key discriminators that drive toward specific levels of business performance.

23. The method of claim 22 further comprising the step of obtaining the data set of business performance questions and numerical responses via a questionnaire containing a plurality of questions that probe into specific performance areas of the business entities.

24. The method of claim 22 further comprising the step of representing the subset of questions and responses as a classification tree wherein a first question and response from the subset of questions and responses is identified as a root node and all other questions and responses from the subset of questions and responses are identified as sub-nodes.

25. The method of claim 24 wherein the rule induction technique applied to the questions and responses comprises a CART algorithm.

26. The method of claim 22 wherein the business performance questions and numerical responses are stored in a database as separate company profiles to define the knowledge-base from which a rule induction engine may identify the key discriminators of business performance.

27. A benchmarking system to rank and position a the performance of an entity in specific areas by identifying key discriminators that drive toward specific levels of business performance comprising means for obtaining business performance information for a plurality of entities, the business performance information including a first set of performance criteria, each performance criterion having an associated value for scoring the performance of the plurality of entities, means for storing the business performance information and a rule induction engine for evaluating the first set of performance criteria and associated values to identify an optimal subset of performance criteria that are the key discriminators.

28. The system of claim 27 wherein the means for obtaining business performance information comprises a questionnaire.

29. The system of claim 27 wherein the rule induction engine comprises a CBR algorithm.

30. The system of claim 29 wherein the CBR algorithm is a CART algorithm.

31. The system of claim 27 further comprising spreadsheet application and presentation application for generating reports based the business information data.

32. The system of claim 31 wherein the spreadsheet application supplies the business performance information to the rule induction engine.