FRAMEWORK FOR MANAGING FEATURES ACROSS ENVIRONMENTS

• US 20190325085A1
• Filed: 04/20/2018
• Published: 10/24/2019
• Est. Priority Date: 04/20/2018
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Stein, David J., Ogilvie, Paul T., Chen, Bee-Chung, Chatterjee, Shaunak, Gariba, Priyanka, Wu, Ke, Tang, Grace W., Luo, Yangchun, Chen, Boyi, Yadav, Amit, Wang, Ruoyang, Gadde, Divya, Gao, Wenxuan, Chandak, Amit, Agnihotri, Varnit, Zhuang, Wei, Young, Joel D., Zhang, Weidong
• Granted Patent: US 11,704,370 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/907 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06F 9/44505 Configuring for program initiating, e.g. using registry, configuration files
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

The disclosed embodiments provide a system for processing data. During operation, the system obtains a feature configuration for a feature. Next, the system obtains, from the feature configuration, an anchor containing metadata for accessing the feature in an environment. The system then uses one or more attributes of the anchor to retrieve one or more feature values of the feature from the environment. Finally, the system provides the one or more feature values for use with one or more machine-learning models.

Claims

1. A method, comprising:

obtaining a feature configuration for a feature;

obtaining, by a computer system from the feature configuration, an anchor comprising metadata for accessing the feature in an environment;

using one or more attributes of the anchor to retrieve, by the computer system, one or more feature values of the feature from the environment; and

providing the one or more feature values for use with one or more machine learning models.

2. The method of claim 1, wherein obtaining the feature configuration for the feature comprises:

obtaining a feature name for the feature from a request for feature values; and

using a global namespace of features across multiple environments to match the feature name to the feature configuration.

3. The method of claim 1, wherein using the one or more attributes of the anchor to retrieve the one or more feature values from the environment comprises:

obtaining an entity key from the request; and

using the entity key and the one or more attributes to obtain the one or more feature values from the environment.

4. The method of claim 3, wherein using the entity key and the one or more attributes to obtain the one or more feature values from the environment comprises:

using the one or more attributes to generate a batch request comprising the entity key and additional entity keys from other requests; and

using the batch request to retrieve a set of feature values for the feature from the environment.

5. The method of claim 1, wherein the feature configuration further comprises:

the feature name;

a feature type of the feature; and

an entity domain of the feature.

6. The method of claim 1, further comprising:

obtaining a join configuration comprising the feature and one or more additional features; and

using the join configuration to join the one or more feature values and additional feature values for the one or more additional features with observation data associated with the one or more feature values and the additional feature values.

7. The method of claim 6, wherein using the join configuration to join the feature and the one or more additional features with the observation data comprises:

obtaining one or more join keys from the join configuration; and

using the one or more join keys to join the feature and the one or more additional features with the observation data.

8. The method of claim 1, wherein using the one or more elements of the anchor to retrieve the one or more features values from the environment comprises:

obtaining a topic identifying a stream of messages from the anchor; and

obtaining, from the environment, the one or more feature values from one or more messages in the stream.

9. The method of claim 1, wherein the one or more attributes comprise at least one of:

a source of the feature within the environment; and

a mechanism for obtaining data related to the feature; and

an operation for producing the one or more feature values from the data.

10. The method of claim 1, wherein the environment is at least one of:

an online environment;

a nearline environment;

an offline environment;

a stream-processing environment;

a search-based environment.

11. The method of claim 1, wherein the environment is contained within another environment.

12. A system, comprising:

one or more processors; and

memory storing instructions that, when executed by the one or more processors, cause the system to;

obtain a feature configuration for a feature;

obtain, from the feature configuration, an anchor comprising metadata for accessing the feature in an environment;

use one or more attributes of the anchor to retrieve one or more feature values of the feature from the environment; and

provide the one or more feature values for use with one or more machine learning models.

13. The system of claim 12, wherein using the one or more attributes of the anchor to retrieve the one or more feature values from the environment comprises:

obtaining an entity key from the request; and

using the entity key and the one or more attributes to obtain the one or more feature values from the environment.

14. The system of claim 13, wherein using the entity key and the one or more attributes to obtain the one or more feature values from the environment comprises:

using the one or more attributes to generate a batch request comprising the entity key and additional entity keys from other requests; and

using the batch request to retrieve a set of feature values for the feature from the environment.

15. The system of claim 12, wherein the feature configuration further comprises:

the feature name;

a feature type of the feature; and

an entity domain of the feature.

16. The system of claim 12, wherein using the one or more elements of the anchor to retrieve the one or more features values from the environment comprises:

obtaining a topic identifying a stream of messages from the anchor; and

obtaining, from the environment, the one or more feature values from one or more messages in the stream.

17. The system of claim 12, wherein the one or more attributes comprise at least one of:

a source of the feature within the environment; and

a mechanism for obtaining data related to the feature; and

an operation for producing the one or more feature values from the data.

18. The system of claim 12, wherein the environment is at least one of:

an online environment;

a nearline environment;

an offline environment;

a stream-processing environment;

a search-based environment.

19. A non-transitory computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method, the method comprising:

obtaining a feature configuration for a feature;

obtaining, from the feature configuration, an anchor comprising metadata for accessing the feature in an environment;

using one or more attributes of the anchor to retrieve one or more feature values of the feature from the environment; and

providing the one or more feature values for use with one or more machine learning models.

20. The non-transitory computer-readable storage medium of claim 19, wherein the method further comprises:

obtaining a join configuration comprising the feature and one or more additional features; and

using the join configuration to join the one or more feature values and additional feature values for the one or more additional features with observation data associated with the one or more feature values and the additional feature values.