Staged application programming interface

• US 10,417,059 B1
• Filed: 08/03/2018
• Issued: 09/17/2019
• Est. Priority Date: 08/03/2018
• Status: Active Grant

• Current Assignee: Intuit, Inc.
• Original Assignee: Intuit, Inc.
• Inventors: Arya, Amit, Kunjachan, George Chiramattel, Vogel, Peter, Riding, Tyler, Martin, Jennifer
• Primary Examiner(s): Dorais, Craig C
• Application Number:
  US16/054,729
• Time in Patent Office: 410 Days
• Field of Search: None
• CPC Class Codes: G06F 16/211 Schema design and management
• CPC Class Codes: G06F 16/2379 Updates performed during online database operations; commit processing
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06F 9/54 Interprogram communication

Abstract

A method may include receiving, via a network and from a source application, partial data including values for parameters defined according to an application programming interface (API). Each parameter may be defined by the API as optional. The partial data may be incomplete for use by a target application. The method may further include obtaining a set of contextual data correlated with the partial data, generating augmented partial data by combining the partial data and contextual data selected from the set of contextual data, determining that the augmented partial data satisfies a transaction completion criterion of a transaction schema, and in response to determining that the augmented partial data satisfies the transaction completion criterion, sending the augmented partial data to a transaction store used by the target application.

Claims

1. A method, comprising:

receiving, via a network and from a source application, partial data comprising values for a plurality of parameters defined according to an application programming interface (API), wherein each of the plurality of parameters is defined by the API as optional, wherein the partial data is incomplete for use by a target application;

obtaining a first set of contextual data correlated with the partial data;

generating augmented partial data by combining the partial data and first contextual data selected from the first set of contextual data;

determining that the augmented partial data satisfies a transaction completion criterion of a transaction schema; and

in response to determining that the augmented partial data satisfies the transaction completion criterion, sending the augmented partial data to a transaction store used by the target application.

2. The method of claim 1, further comprising:

assigning the first contextual data to a parameter of the plurality of parameters without a corresponding value in the partial data.

3. The method of claim 1, further comprising:

assigning, using a first ranking model, a first plurality of confidence levels to the first set of contextual data, wherein the first contextual data is selected according to the first plurality of confidence levels;

in response to determining that no confidence level in the first plurality of confidence levels exceeds a threshold, obtaining a second set of contextual data correlated with the partial data;

assigning, using the first ranking model, a second plurality of confidence levels to the second set of contextual data;

selecting second contextual data according to the second plurality of confidence levels assigned to the second set of contextual data; and

determining that the confidence level assigned to the second contextual data exceeds the threshold,

wherein generating the augmented partial data comprises combining the second contextual data with the partial data after determining that the first contextual data is not assigned a confidence level satisfying the threshold.

4. The method of claim 3, further comprising:

identifying a first context store using a first rule based on a user profile, wherein the first set of contextual data is received from the first context store in response to sending a first request for data correlated with the partial data; and

identifying a second context store using a second rule based on the user profile, wherein the second set of contextual data is received from the second context store in response to sending a second request for data correlated with the partial data.

5. The method of claim 4, wherein obtaining the first set of contextual data comprises matching, using a predefined matching criterion, the partial data with contextual data of the first context store.

6. The method of claim 1, further comprising:

extracting, from the transaction store, user-specific training data comprising transaction data for a user; and

training the first ranking model using the user-specific training data.

7. The method of claim 1, further comprising:

assigning, using a first ranking model, a first plurality of confidence levels to the first set of contextual data, wherein the first contextual data is selected according to the first plurality of confidence levels; and

in response to determining that no confidence level in the first plurality of confidence levels exceeds a threshold;

extracting, from the transaction store, user-generic training data comprising transaction data for a group of users;

training a second ranking model using the user-generic training data;

assigning, using the second ranking model, a second plurality of confidence levels to the first set of contextual data;

selecting second contextual data according to the second plurality of confidence levels assigned to the first set of contextual data; and

determining that the confidence level assigned to the second contextual data exceeds the threshold,

wherein generating the augmented partial data comprises combining the second contextual data with the partial data after determining that the first contextual data is not assigned a confidence level satisfying the threshold.

8. A system, comprising:

a memory coupled to a processor;

a repository comprising augmented partial data, a first ranking model, and a transaction schema comprising a transaction completion criterion;

an application programming interface (API) that defines a plurality of parameters, wherein each of the plurality of parameters is defined as optional;

a stage controller executing on the processor and using the memory, configured to;

receive, via a network and from a source application, partial data comprising values for one or more of the plurality of parameters, wherein the partial data is incomplete for use by a target application;

generate the augmented partial data by combining the partial data and first contextual data selected from a first set of contextual data correlated with the partial data;

determine that the augmented partial data satisfies the transaction completion criterion; and

in response to determining that the augmented partial data satisfies the transaction completion criterion, send the augmented partial data to a transaction store used by the target application; and

a contextualizer executing on the processor and using the memory, configured to;

obtain the first set of contextual data.

9. The system of claim 8, wherein the stage controller is further configured to:

assign the first contextual data to a parameter of the plurality of parameters without a corresponding value in the partial data.

10. The system of claim 8,

wherein the system further comprises a first ranking model, and

wherein the contextualizer is further configured to:

assign, using the first ranking model, a first plurality of confidence levels to the first set of contextual data, wherein the first contextual data is selected according to the first plurality of confidence levels; and

in response to determining that no confidence level in the first plurality of confidence levels exceeds a threshold;

obtain a second set of contextual data correlated with the partial data;

assign, using the first ranking model, a second plurality of confidence levels to the second set of contextual data;

select second contextual data according to the second plurality of confidence levels assigned to the second set of contextual data; and

determine that the confidence level assigned to the second contextual data exceeds the threshold,

wherein generating the augmented partial data comprises combining the second contextual data with the partial data after determining that the first contextual data is not assigned a confidence level satisfying the threshold.

11. The system of claim 10, wherein the contextualizer is further configured to:

identify a first context store using a first rule based on a user profile, wherein the first set of contextual data is received from the first context store in response to sending a first request for data correlated with the partial data; and

identify a second context store using a second rule based on the user profile, wherein the second set of contextual data is received from the second context store in response to sending a second request for data correlated with the partial data.

12. The system of claim 11, wherein the contextualizer is further configured to obtain the first set of contextual data by matching, using a predefined matching criterion, the partial data with contextual data of the first context store.

13. The system of claim 8, wherein the contextualizer is further configured to:

extract, from the transaction store, user-specific training data comprising transaction data for a user; and

train the first ranking model using the user-specific training data.

14. The system of claim 8,

wherein the repository further comprises a first ranking model and a second ranking model, and

wherein the contextualizer is further configured to:

assign, using the first ranking model, a first plurality of confidence levels to the first set of contextual data, wherein the first contextual data is selected according to the first plurality of confidence levels; and

in response to determining that no confidence level in the first plurality of confidence levels exceeds a threshold;

extract, from the transaction store, user-generic training data comprising transaction data for a group of users;

train the second ranking model using the user-generic training data;

assign, using the second ranking model, a second plurality of confidence levels to the first set of contextual data;

select second contextual data according to the second plurality of confidence levels assigned to the first set of contextual data; and

determine that the confidence level assigned to the second contextual data exceeds the threshold,

wherein generating the augmented partial data comprises combining the second contextual data with the partial data after determining that the first contextual data is not assigned a confidence level satisfying the threshold.

15. An application programming interface (API), comprising executable code stored in a memory and executed on a processor, further comprising instructions that perform:

receiving, via a network and from a source application, partial data comprising values for one or more of a plurality of parameters, wherein each of the plurality of parameters is defined as optional, wherein the partial data is incomplete for use by a target application; and

transmitting, to a staging platform executing on the processor and using the memory, the partial data, wherein the staging platform;

obtains a first set of contextual data correlated with the partial data;

generates augmented partial data by combining the partial data and first contextual data selected from the first set of contextual data;

determines that the augmented partial data satisfies a transaction completion criterion of a transaction schema; and

in response to determining that the augmented partial data satisfies the transaction completion criterion, sends the augmented partial data to a transaction store used by the target application.

16. The API of claim 15, further comprising instructions that perform:

assigning the first contextual data to a parameter of the plurality of parameters without a corresponding value in the partial data.

17. The API of claim 15, wherein the staging platform further:

assigns, using a first ranking model, a first plurality of confidence levels to the first set of contextual data, wherein the first contextual data is selected according to the first plurality of confidence levels;

in response to determining that no confidence level in the first plurality of confidence levels exceeds a threshold, obtains a second set of contextual data correlated with the partial data;

assigns, using the first ranking model, a second plurality of confidence levels to the second set of contextual data;

selects second contextual data according to the second plurality of confidence levels assigned to the second set of contextual data; and

determines that the confidence level assigned to the second contextual data exceeds the threshold,

wherein generating the augmented partial data comprises combining the second contextual data with the partial data after determining that the first contextual data is not assigned a confidence level satisfying the threshold.

18. The API of claim 17, wherein the staging platform further:

identifies a first context store using a first rule based on a user profile, wherein the first set of contextual data is received from the first context store in response to sending a first request for data correlated with the partial data; and

identifies a second context store using a second rule based on the user profile, wherein the second set of contextual data is received from the second context store in response to sending a second request for data correlated with the partial data.

19. The API of claim 18, wherein the staging platform further obtains the first set of contextual data by matching, using a predefined matching criterion, the partial data with contextual data of the first context store.

20. The API of claim 15, wherein the staging platform further:

extracts, from the transaction store, user-specific training data comprising transaction data for a user; and

trains the first ranking model using the user-specific training data.