Cloud sharing and selection of machine learning models for service use

Patent metadata

Field	Value
Publication number	US-11222281-B2
Application number	US-201816018284-A
Country	US
Kind code	B2
Filing date	Jun 26, 2018
Priority date	Jun 26, 2018
Publication date	Jan 11, 2022
Grant date	Jan 11, 2022

Abstract

An approach is provided in which an information handling system performs multiple tests using a cognitive service and multiple trained machine learning models on user data corresponding to a user application. For each of the multiple tests, a different one of the trained machine learning models is utilized. The information handling system generates results from the tests and then selects at least one of the trained machine learning models based on the test results. In turn, the information handling system assigns the cognitive service and the selected trained machine learning models to the user application.

First claim

The invention claimed is:

1. A method implemented by an information handling system that includes a memory and a processor, the method comprising:
receiving, from a user system, a set of user sample data corresponding to a user application executing on the user system;
performing a plurality of tests on a plurality of cognitive services and a plurality of pre-trained machine learning models by inputting the user sample data into each of the plurality of cognitive services and each of the plurality of pre-trained machine learning models, wherein a different one of the plurality of pre-trained machine learning models is utilized during each one of the plurality of tests;
selecting one of the plurality of cognitive services and at least one of the plurality of pre-trained machine learning models based, at least in part, on a plurality of test results generated from the plurality of tests;
binding the selected at least one pre-trained machine learning model to the selected cognitive service; and
in response to the binding, processing a set of user runtime data from the user system using the selected cognitive service and the selected at least one pre-trained machine learning model.

2. The method of claim 1 further comprising:
prior to performing the plurality of tests on the user sample data:
determining a data type of the user sample data; and
selecting a subset of a plurality of cognitive services based on the data type; and

in response to selecting the subset of the plurality of cognitive services:
performing the plurality of tests on the user sample data using each of the subset of cognitive services and the plurality of pre-trained machine learning models.

3. The method of claim 1 further comprising:
determining to assign a different pre-trained machine learning model to the cognitive service based on analyzing a set of results from the processing of the user runtime data; and
dynamically binding the different pre-trained machine learning model to the cognitive service.

4. The method of claim 1 wherein each of the plurality of pre-trained machine learning models were trained prior to the performing of the plurality of tests and are unmodified during the performing of the plurality of tests, the selecting, and the binding.

5. The method of claim 1 further comprising:
receiving the plurality of pre-trained machine learning models from a set of expert providers;
storing the plurality of pre-trained machine learning models in an open catalog repository; and
selecting the plurality of pre-trained machine learning models from the open catalog repository to perform the plurality of tests.

6. The method of claim 1 wherein the user sample data comprises a set of natural language queries.

7. An information handling system comprising:
one or more processors;
a memory coupled to at least one of the processors;
a set of computer program instructions stored in the memory and executed by at least one of the processors in order to perform actions of:
receiving, from a user system, a set of user sample data corresponding to a user application executing on the user system
performing a plurality of tests on a plurality of cognitive services and a plurality of pre-trained machine learning models by inputting the user sample data into each of the plurality of cognitive services and each of the plurality of pre-trained machine learning models, wherein a different one of the plurality of pre-trained machine learning models is utilized during each one of the plurality of tests;
selecting one of the plurality of cognitive services and at least one of the plurality of pre-trained machine learning models based, at least in part, on a plurality of test results generated from the plurality of tests;
binding the selected at least one pre-trained machine learning model to the selected cognitive service; and
in response to the binding, processing a set of user runtime data from the user system using the selected cognitive service and the selected at least one pre-trained machine learning model.

8. The information handling system of claim 7 wherein the processors perform additional actions comprising:
prior to performing the plurality of tests on the user sample data:
determining a data type of the user sample data; and
selecting a subset of a plurality of cognitive services based on the data type; and

in response to selecting the subset of the plurality of cognitive services:
performing the plurality of tests on the user sample data using each of the subset of cognitive services and the plurality of pre-trained machine learning models.

9. The information handling system of claim 7 wherein the processors perform additional actions comprising:
determining to assign a different pre-trained machine learning model to the cognitive service based on analyzing a set of results from the processing of the user runtime data; and
dynamically binding the different pre-trained machine learning model to the cognitive service.

10. The information handling system of claim 7 wherein each of the plurality of pre-trained machine learning models were trained prior to the performing of the plurality of tests and are unmodified during the performing of the plurality of tests, the selecting, and the binding.

11. The information handling system of claim 7 wherein the processors perform additional actions comprising:
receiving the plurality of pre-trained machine learning models from a set of expert providers;
storing the plurality of pre-trained machine learning models in an open catalog repository; and
selecting the plurality of pre-trained machine learning models from the open catalog repository to perform the plurality of tests.

12. The information handling system of claim 7 wherein the user sample data comprises a set of natural language queries.

13. A computer program product stored in a computer readable storage medium, comprising computer program code that, when executed by an information handling system, causes the information handling system to perform actions comprising:
receiving, from a user system, a set of user sample data corresponding to a user application executing on the user system;
performing a plurality of tests on a plurality of cognitive services and a plurality of pre-trained machine learning models by inputting the user sample data into each of the plurality of cognitive services and each of the plurality of pre-trained machine learning models, wherein a different one of the plurality of pre-trained machine learning models is utilized during each one of the plurality of tests;
selecting one of the plurality of cognitive services and at least one of the plurality of pre-trained machine learning models based, at least in part, on a plurality of test results generated from the plurality of tests;
binding the selected at least one pre-trained machine learning model to the selected cognitive service; and
in response to the binding, processing a set of user runtime data from the user system using the selected cognitive service and the selected at least one pre-trained machine learning model.

14. The computer program product of claim 13 wherein the information handling system performs further actions comprising:
prior to performing the plurality of tests on the user sample data:
determining a data type of the user sample data; and
selecting a subset of a plurality of cognitive services based on the data type; and

in response to selecting the subset of the plurality of cognitive services:
performing the plurality of tests on the user sample data using each of the subset of cognitive services and the

Assignees

• IBM

Inventors

• Kozhaya Joseph N
• Allen Corville O
• Freed Andrew R

Classifications

• G06N20/00 Primary

  Machine learning · CPC title

• G06F9/44521

  Dynamic linking or loading; Link editing at or after load time, e.g. Java class loading · CPC title

• G06F16/3329 Primary

  Natural language query formulation · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06F8/36

  Software reuse · CPC title