Search and exploration using analytics reference model

• US 8,145,615 B2
• Filed: 11/26/2008
• Issued: 03/27/2012
• Est. Priority Date: 11/26/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Mital, Vijay, Green, David G., Rubin, Darryl E.
• Primary Examiner(s): Pham, Hung Q
• Application Number:
  US12/324,480
• Publication Number:
  US 20100131546A1
• Time in Patent Office: 1,217 Days
• Field of Search: 707/705, 707/777
• US Class Current: 707/705
• CPC Class Codes: G06F 16/29 Geographical information databases

Abstract

Searching and exploration using a data-driven analytics model. The analytics model includes an analytical modeling component that defines analytical relationships between model variables using a number of analytical relations. In response to a search request, the output variable(s) of the solve operation are identified. The output variable(s) may have even been identified based on the search request. The analytical relations of the model may then be used to solve for the identified output variable(s). The resulting value(s) for the now solved-for output variable(s) may then be used to formulate the response to the search request. The nature of the response may vary depending on the scope of the application that embodied the search request capability. The results of the search request may be used for further exploration of the model by, for example, submitting follow-up search requests, resulting in follow-up solve operations.

Claims

1. A method for searching using analytics reference model, the method comprising the steps of:

instantiating, by a computing system, an analytical modeling component that defines analytical relationships between a plurality of model variables using a plurality of analytical relations, wherein the analytical modeling component uses the plurality of analytical relations to identify which of the plurality of model variables is or are to be solved for, and which of the plurality of model variables is or are an input variable, and wherein an identity of any output model variables and an identity of any of the input model variables may change from one solve operation to the next;

accessing a received first search request which requests data related to at least one category of a plurality of data categories, the received first search request also providing at least one input variable;

identifying at least one output variable that should be solved for in response to the received first search request, wherein the identified at least one output variable corresponds to the provided at least one input variable, wherein the identifying including;

navigating a data taxonomy which taxonomizes a plurality of data items into a corresponding one of the plurality of related data categories, to identify at least one data item related to the requested at least one category corresponding to the received first search request;

binding the identified at least one data item to the identified at least one output variable using the navigated data taxonomy;

using (i) a plurality of analytical relations between the provided at least one input variable and the identified at least one output variable and (ii) the at least one data item identified as part of the data taxonomy to solve for at least one value of the identified at least one output variable; and

formulating a response to the received first search request using the at least one solved value of the identified at least one output variable.

2. The method in accordance with claim 1, wherein the search request is a first search request, the method further comprising:

accessing a received second search request;

using the received second search request to identify at least one output variable that should be solved for in order to respond to the received second search request;

using the plurality of analytical relations and the data taxonomy to solve for at least one value of the identified at least one output variable that was identified using the received second search request; and

formulating a response to the received second search request using the at least one solved value of the at least one output variable that was identified using the received second search request.

3. The method in accordance with claim 2, further comprising:

accessing a received third search request;

using the received second search request to identify at least one output variable that should be solved for in order to respond to the received third search request;

using the plurality of analytical relations and the data taxonomy to solve for at least one value of the identified at least one output variable that was identified using the received third search request; and

formulating a response to the received third search request using the at least one solved value of the at least one output variable that was identified using the received third search request.

4. The method in accordance with claim 3, wherein the identity of the at least one output variable identified using the received third search request is different than the identity of the at least one output variable identified using the received first search request and is different than the identity of the at least one output variable identified using the received second search request.

5. The method in accordance with claim 2, wherein the identity of the at least one output variable identified using the received second search request is different than the identity of the at least one output variable identified using the received first search request.

6. The method in accordance with claim 2, wherein the identity of the at least one output variable identified using the received second search request is the same as the identity of the at least one output variable identified using the received first search request.

7. The method in accordance with claim 2, wherein the at least one solved value used to formulate a response to the received first request are used as at least one input variable when using the plurality of analytical relations and the data taxonomy to solve for at least one value of the identified at least one output variable that was identified using the received second search request.

8. The method in accordance with claim 2, wherein using the plurality of analytical relations and the data taxonomy to solve for at least one value of the identified at least one output variable that was identified using the received second search request comprises re-using at least one value previously solved for when solving for at least one value of the at least one output variable of the received first search request.

9. The method in accordance with claim 1, wherein at least some of the plurality of analytical relations are equations.

10. The method in accordance with claim 1, wherein at least some of the plurality of analytical relations are constraints.

11. The method in accordance with claim 1, wherein at least some of the plurality of analytical relations are rules.

12. The method in accordance with claim 1, wherein at least some of the plurality of analytical relations are simulations.

13. The method in accordance with claim 12, wherein using (i) the plurality of analytical relations between the at least one input variable and the at least one output variable and (ii) the at least one data item identified as part of the data taxonomy to solve for at least one value of the identified at least one output variable comprises performing at least one simulation specified in the plurality of analytical relations.

14. The method in accordance with claim 1, wherein the method is performed by the computing system without contemporaneous assistance from a human being.

15. The method in accordance with claim 1, the method further comprising:

accessing a data source that is a source for the plurality of data items when navigating the data taxonomy; and

utilizing a data-model binding component configured to bind portions of the plurality of data items to corresponding model parameters of the plurality of model parameters when binding the at least one data item to at least one output variable.

16. The method in accordance with claim 1, the method further comprising:

accessing a view components repository that contains a plurality of heterogenic view components, each corresponding to a visual item that may be displayed, at least some of which being parameterized;

using a model-visual binding component to bind the plurality of model parameter values one or more parameters of at least one parameterized view component contained within the view components repository; and

using a view composition module to formulate instructions for rendering a view that contains at least some of the visual items corresponding to the plurality of heterogenic view components.

17. The method in accordance with claim 1, wherein the search request requires a noise calculation in order to formulate a response to the received first search request.

18. The method in accordance with claim 1, wherein the received first search request specifies a range for a value of a model variable that is not yet solved,where the not-yet solved for variable becomes an output model variable for a solve that is performed in order to satisfy the search received first search request.

19. The method in accordance with claim 1, the method further comprising:

receiving a second search request;

identifying at least one output variable of the plurality of model variables that should be solved for in order to respond to the second search request;

using the plurality of analytical relations and the at least one solved value of the at least one output variable identified using the first search request to then solve for at least one other output variable identified using the second search request; and

formulating a response to the second search request using the at least one solved value of the at least one other output variable identified using the second search request.

20. A computer program product comprised of a physical storage media which contains executable instructions executed by at least one processor to perform a method for searching using analytics reference model comprising the steps of:

instantiating an analytical modeling component that defines analytical relationships between a plurality of model variables using a plurality of analytical relations, wherein the analytical modeling component uses the plurality of analytical relations to identify which of the plurality of model variables is or are to be solved for, and which of the plurality of model variables is or are an input variable, and wherein an identity of any output model variables and an identity of any of the input model variables may change from one solve operation to the next;

accessing a received first search request which requests data related to at least one category of a plurality of data categories, the received first search request also providing at least one input variable;

identifying at least one output variable that should be solved for in response to the received first search request, wherein the identified at least one output variable corresponds to the provided at least one input variable, wherein the identifying including;

navigating a data taxonomy which taxonomizes a plurality of data items into a corresponding one of the plurality of related data categories, to identify at least one data item related to the requested at least one category corresponding to the received first search request;

binding the identified at least one data item to the identified at least one output variable using the navigated data taxonomy;

using (i) a plurality of analytical relations between the provided at least one input variable and the identified at least one output variable and (ii) the at least one data item identified as part of the data taxonomy to solve for at least one value of the identified at least one output variable; and

formulating a response to the received first search request using the at least one solved value of the identified at least one output variable.