Identifying resource allocation discrepancies

• US 10,482,407 B2
• Filed: 11/14/2016
• Issued: 11/19/2019
• Est. Priority Date: 11/14/2016
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: Apptio Inc (International Business Machines Corporation)
• Inventors: Tijerina, David, Krishnan, Venkatesh
• Primary Examiner(s): Misiaszek, Amber A
• Application Number:
  US15/351,313
• Publication Number:
  US 20180137445A1
• Time in Patent Office: 1,100 Days
• Field of Search: 705 711- 742
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06Q 10/06313 Resource planning in a project environment

Abstract

Embodiments are directed to identifying allocation discrepancies. Data models and Benchmark models may be provided to an analysis engine. Discrepancy models may be provided to the analysis engine, such that each discrepancy model may be arranged to include one or more rules. The analysis engine may be employed to search for discrepancies in the data models based on the discrepancy models and the benchmark models. If discrepancies may be identified by the analysis engine, one or more notifications may be provided to one or more users. Also, the discrepancy models may be modified based on subsequent feedback provided by the one or more users. Correlations in the resource allocation values may be identified based on machine learning that includes one or more of linear regression, deep learning neural networks, or the like. And, additional discrepancy models may be provided based on the identified correlations.

Claims

1. A method for identifying allocation discrepancies using a network computer that employs one or more processors to execute instructions that perform actions, comprising:

instantiating a resource modeling engine to perform actions, including;

providing a plurality of data models and a plurality of benchmark models to a separately instantiated analysis engine, wherein the plurality of data models are each comprised of two or more nodes and one or more edges from a memory that is configured and arranged to store the plurality of data models, and wherein each node represents a model object and each edge represents a resource allocation; and

providing one or more discrepancy models to the analysis engine, wherein each discrepancy model includes one or more rules, wherein the analysis engine is employed to search for discrepancies in the plurality of data models based on the one or more discrepancy models and the plurality of benchmark models; and

instantiating a visualization engine to use one or more discrepancies that are identified by the analysis engine to perform further actions, including;

providing one or more notifications to one or more users, wherein geolocation information based on electronic signals processed by a Global Positioning System (GPS) device is employed to improve understanding of a visual display for one or more users by modifying the visual display of the notifications based on a location of a client computer employed by the one or more users, wherein the modification includes one or more of time zone, currency, calendar format, or language and, wherein the resource modeling engine modifies the one or more discrepancy models based on subsequent feedback provided by the one or more users in the visual display and also employs the one or more of time zone, currency, calendar format, or language to modify one or more of a user interface, a report, an internal process, or a database.

2. The method of claim 1, wherein the one or more rules include one or more threshold conditions that correspond to a discrepancy in the plurality of data models.

3. The method of claim 1, further comprising:

employing the analysis engine to traverse one or more data models of the plurality of data models;

providing one or more model objects based on the traversal of the one or more data models; and

identifying the one or more discrepancies based on an execution of one or more rules, wherein the execution compares one or more threshold values in the one or more rules to resource allocation values associated with the one or more model objects.

4. The method of claim 1, further comprising:

traversing one or more data models associated with one or more customers;

traversing one or more other data models associated with one or more other customers;

executing the one or more rules to compare one or more model objects from the one or more data models and one or more other model objects from the one or more other data models; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

5. The method of claim 1, further comprising:

traversing one or more data models associated with one or more industry segments to provide one or more model objects;

executing one or more rules to compare the one or more model objects, wherein the one or more compared model objects are associated with separate industry segments; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

6. The method of claim 1, wherein identifying the one or more discrepancies based on the execution of the one or more rules, further comprises, comparing one or more resource allocation values associated with one or more model objects in a data model with a mean resource allocation value that is associated with one or more corresponding benchmark objects.

7. The method of claim 1, further comprising:

traversing the plurality of data models to identify clusters of resource allocation values associated with the one or more model objects;

identifying one or more correlations in the resource allocation values based on machine learning that includes one or more of linear regression, or deep learning neural networks; and

providing one or more additional discrepancy models based on the one or more identified correlations.

8. A processor readable non-transitory storage media that includes instructions for identifying allocation discrepancies, wherein execution of the instructions by one or more hardware processors performs actions, comprising:

instantiating a resource modeling engine to perform actions, including;

providing a plurality of data models and a plurality of benchmark models to a separately instantiated analysis engine, wherein the plurality of data models are each comprised of two or more nodes and one or more edges from a memory that is configured and arranged to store the plurality of data models, and wherein each node represents a model object and each edge represents a resource allocation; and

providing one or more discrepancy models to the analysis engine, wherein each discrepancy model includes one or more rules, wherein the analysis engine is employed to search for discrepancies in the plurality of data models based on the one or more discrepancy models and the plurality of benchmark models; and

instantiating a visualization engine to use one or more discrepancies that are identified by the analysis engine to perform further actions, including;

providing one or more notifications to one or more users, wherein geolocation information based on electronic signals processed by a Global Positioning System (GPS) device is employed to improve understanding of a visual display for one or more users by modifying the visual display of the notifications based on a location of a client computer employed by the one or more users, wherein the modification includes one or more of time zone, currency, calendar format, or language and, wherein the resource modeling engine modifies the one or more discrepancy models based on subsequent feedback provided by the one or more users in the visual display, wherein the resource modeling engine modifies the one or more discrepancy models based on subsequent feedback provided by the one or more users in the visual display and also employs the one or more of time zone, currency, calendar format, or language to modify one or more of a user interface, a report, an internal process, or a database.

9. The media of claim 8, wherein the one or more rules include one or more threshold conditions that correspond to a discrepancy in the plurality of data models.

10. The media of claim 8, further comprising:

employing the analysis engine to traverse one or more data models of the plurality of data models;

providing one or more model objects based on the traversal of the one or more data models; and

identifying the one or more discrepancies based on an execution of one or more rules, wherein the execution compares one or more threshold values in the one or more rules to resource allocation values associated with the one or more model objects.

11. The media of claim 8, further comprising:

traversing one or more data models associated with one or more customers;

traversing one or more other data models associated with one or more other customers;

executing the one or more rules to compare one or more model objects from the one or more data models and one or more other model objects from the one or more other data models; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

12. The media of claim 8, further comprising:

traversing one or more data models associated with one or more industry segments to provide one or more model objects;

executing one or more rules to compare the one or more model objects, wherein the one or more compared model objects are associated with separate industry segments; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

13. The media of claim 8, wherein identifying the one or more discrepancies based on the execution of the one or more rules, further comprises, comparing one or more resource allocation values associated with one or more model objects in a data model with a mean resource allocation value that is associated with one or more corresponding benchmark objects.

14. The media of claim 8, further comprising:

traversing the plurality of data models to identify clusters of resource allocation values associated with the one or more model objects;

identifying one or more correlations in the resource allocation values based on machine learning that includes one or more of linear regression, or deep learning neural networks; and

providing one or more additional discrepancy models based on the one or more identified correlations.

15. A network computer for identifying allocation discrepancies, comprising:

a transceiver that communicates over the network;

a memory that stores at least instructions; and

one or more processor devices that execute instructions that perform actions, including;

instantiating a resource modeling engine to perform actions, including;

providing a plurality of data models and a plurality of benchmark models to a separately instantiated analysis engine, wherein the plurality of data models are each comprised of two or more nodes and one or more edges from a memory that is configured and arranged to store the plurality of data models, and wherein each node represents a model object and each edge represents a resource allocation; and

providing one or more discrepancy models to the analysis engine, wherein each discrepancy model includes one or more rules, wherein the analysis engine is employed to search for discrepancies in the plurality of data models based on the one or more discrepancy models and the plurality of benchmark models; and

instantiating a visualization engine to use one or more discrepancies that are identified by the analysis engine to perform further actions, including;

providing one or more notifications to one or more users, wherein geolocation information based on electronic signals processed by a Global Positioning System (GPS) device is employed to improve understanding of a visual display for one or more users by modifying the visual display of the notifications based on a location of a client computer employed by the one or more users, wherein the modification includes one or more of time zone, currency, calendar format, or language and, wherein the resource modeling engine modifies the one or more discrepancy models based on subsequent feedback provided by the one or more users in the visual display and also employs the one or more of time zone, currency, calendar format, or language to modify one or more of a user interface, a report, an internal process, or a database.

16. The network computer of claim 15, wherein the one or more rules include one or more threshold conditions that correspond to a discrepancy in the plurality of data models.

17. The network computer of claim 15, further comprising:

employing the analysis engine to traverse one or more data models of the plurality of data models;

providing one or more model objects based on the traversal of the one or more data models; and

identifying the one or more discrepancies based on an execution of one or more rules, wherein the execution compares one or more threshold values in the one or more rules to resource allocation values associated with the one or more model objects.

18. The network computer of claim 15, further comprising:

traversing one or more data models associated with one or more customers;

traversing one or more other data models associated with one or more other customers;

executing the one or more rules to compare one or more model objects from the one or more data models and one or more other model objects from the one or more other data models; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

19. The network computer of claim 15, further comprising:

traversing one or more data models associated with one or more industry segments to provide one or more model objects;

executing one or more rules to compare the one or more model objects, wherein the one or more compared model objects are associated with separate industry segments; and

identifying one or more discrepancies based on one or more affirmative results of the comparison.

20. The network computer of claim 15, wherein identifying the one or more discrepancies based on the execution of the one or more rules, further comprises, comparing one or more resource allocation values associated with one or more model objects in a data model with a mean resource allocation value that is associated with one or more corresponding benchmark objects.

21. The network computer of claim 15, further comprising:

traversing the plurality of data models to identify clusters of resource allocation values associated with the one or more model objects;

identifying one or more correlations in the resource allocation values based on machine learning that includes one or more of linear regression, or deep learning neural networks; and

providing one or more additional discrepancy models based on the one or more identified correlations.