Providing component recommendation using machine learning

1. A computer-implemented method for generating recommended components for an object of a user comprising:

accessing, by a management system, a first plurality of historical component entries of a plurality of different entities, wherein each of the first plurality of historical component entries includes an identifier of a component of an object previously purchased by a respective user and an identifier of a service previously performed on the object at a respective entity from the plurality of different entities after the purchase of the object, the component of the object acquired in conjunction with the service being previously performed on the object;

training, by the management system, a machine learning model using training data that is based at least in part on the identifier of the component of the object and the identifier of the service included in each of the first plurality of historical component entries, the trained machine learning model configured to predict for each of a plurality of predetermined component classifications a likelihood of selection of a component corresponding to one of the plurality of predetermined component classifications;

receiving, by the management system, a request for a service to be performed on the object previously purchased by the user at an entity from the plurality of different entities, the request including attributes of the object that was previously purchased by the user and a specific service for the entity to perform on the object at the entity;

applying, by the management system, the attributes of the object to the trained machine learning model responsive to the request, the trained machine learning model outputting for each of the plurality of predetermined component classifications a prediction of the likelihood of user selection of the component corresponding to the predetermined component classification based on the attributes of the object applied to the trained machine learning model;

determining, by the management system, a recommended set of components for the object that are provided by the entity based on the prediction for each of the plurality of predetermined componenet classications; and

providing, by the management system, the recommended set of components for the object to acquire by the user in conjunction with the specific service that was requested by the user being performed on the object at the entity along with a response to the request for the service.

2. The computer-implemented method of claim 1 , wherein the prediction for each of the plurality of predetermined component classifications is a value indicative of a likelihood of selection of the component corresponding to the predetermined component classification by the user.

3. The computer-implemented method of claim 2 , wherein determining the recommended set of components further comprises:

selecting a subset of the plurality of predetermined component classifications having values greater than a threshold; and

determining for each predetermined component classification included in the subset, a component supplier that is preferred by the entity and at least one component provided by the component supplier that corresponds to the predetermined component classification, wherein the recommended set of components includes the determined at least one component from the preferred component supplier for each predetermined component classification included in the subset.

4. The computer-implemented method of claim 1 , further comprising:

accessing, by the management system, a second plurality of historical component entries of the plurality of different entities, wherein each of the second plurality of historical component entries includes an identifier of a component of an object associated with the second historical component entry but does not include an identifier of a service performed on the object, the component of the object acquired by a respective user from a respective entity from the plurality of different entities after purchase of the object without the service being performed on the object at the respective entity; and

generating, by the management system, the training data by classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries into at least one of the plurality of predetermined component classifications based on the identifier of the component of the object included in each of the first plurality of historical component entries and the identifier of the component included in each second plurality of historical component entries.

5. The computer-implemented method of claim 4 , wherein training the machine learning model comprises:

extracting features from each of the first plurality of historical component entries and each of the second plurality of historical component entries included in the training data, the extracted features including a predetermined component classification assigned to each of the first plurality of historical component entries and each the second plurality of historical component entries, features of the objects included in the first plurality of historical component entries, features of services performed on the objects included in the first plurality of historical component entries, and features of the components included the first plurality of historical component entries and the second plurality of historical component entries, wherein the machine learning model is trained using the extracted features.

6. The computer-implemented method of claim 4 , wherein classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries comprises:

identifying, for each of the first plurality of historical component entries and each of the second plurality of historical component entries, a keyword included in the respective historical component entry;

comparing, for each of the first plurality of historical component entries and each of the second plurality of historical component entries, the keyword with the plurality of predetermined component classifications to identify a match between the keyword and at least one of the plurality of predetermined component classifications; and

classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries with their respective matching predetermined component classification.

7. The computer-implemented method of claim 4 , wherein classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries comprises:

accessing a list of identifiers of components of the plurality of different entities, the list including a mapping of each of the identifiers of components to one or more of the plurality of predetermined component classifications;

comparing the identifier of the component included in each of the first plurality of historical component entries and each of the second plurality of historical component entries to the list, the comparison resulting in a match between identifiers of the components from the first plurality of historical component entries and the second plurality of historical component entries and the identifiers of components in the list; and

classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries with the predetermined component classification based on the match.

8. The computer-implemented method of claim 1 , wherein the machine learning model is trained using Smart Adaptive Recommendations (SAR) algorithm.

9. The computer-implemented method of claim 1 , wherein the management system provides the recommended set of components to a client device of the user a plurality of instances including responsive to the request for service, upon check-in of the user at the entity for the service, and user confirmation to begin servicing the object.

10. The computer-implemented method of claim 1 , wherein the management system provides the recommended set of components to the entity and the entity provides the recommended set of services to a client device of the user a plurality of instances including responsive to the request for service, upon check-in of the user at the entity for the service, and user confirmation to begin servicing the object.

11. The computer-implemented method of claim 1 , wherein the object of the user is an automobile, the recommended set of components are accessories for the automobile, and the attributes of the object includes at least one of a make of the automobile, a model of the automobile, a year of the automobile, engine characteristics of the automobile, a manufacturer's suggested retail price (MSRP) of the automobile, and a market class of the automobile.

12. A non-transitory computer-readable storage medium storing executable code for generating recommended components for an object of a user, the code when executed by a computer processor causes the computer processor to perform steps comprising:

accessing, by a management system, a first plurality of historical component entries of a plurality of different entities, wherein each of the first plurality of historical component entries includes an identifier of a component of an object previously purchased by a respective user and an identifier of a service previously performed on the object at a respective entity from the plurality of different entities after the purchase of the object, the component of the object acquired in conjunction with the service being previously performed on the object;

training, by the management system, a machine learning model using training data that is based at least in part on the identifier of the component of the object and the identifier of the service included in each of the first plurality of historical component entries, the trained machine learning model configured to predict for each of a plurality of predetermined component classifications a likelihood of selection of a component corresponding to one of the plurality of predetermined component classifications;

receiving, by the management system, a request for a service to be performed on the object previously purchased by the user at an entity from the plurality of different entities, the request including attributes of the object that was previously purchased by the user and a specific service for the entity to perform on the object at the entity;

applying, by the management system, the attributes of the object to the trained machine learning model responsive to the request, the trained machine learning model outputting for each of the plurality of predetermined component classifications a prediction of the likelihood of user selection of the component corresponding to the predetermined component classification based on the attributes of the object applied to the trained machine learning model;

determining, by the management system, a recommended set of components for the object that are provided by the entity based on the prediction for each of the plurality of predetermined componenet classications; and

providing, by the management system, the recommended set of components for the object to acquire by the user in conjunction with the specific service that was requested by the user being performed on the object at the entity along with a response to the request for the service.

13. The non-transitory computer-readable storage medium of claim 12 , wherein the prediction for each of the plurality of predetermined component classifications is a value indicative of a likelihood of selection of the component corresponding to the predetermined component classification by the user.

14. The non-transitory computer-readable storage medium of claim 13 , wherein determining the recommended set of components further comprises:

selecting a subset of the plurality of predetermined component classifications having values greater than a threshold; and

determining for each predetermined component classification included in the subset, a component supplier that is preferred by the entity and at least one component provided by the component supplier that corresponds to the predetermined component classification, wherein the recommended set of components includes the determined at least one component from the preferred component supplier for each predetermined component classification included in the subset.

15. The non-transitory computer-readable storage medium of claim 12 , wherein the code when executed by the computer processor causes the computer processor to perform further steps comprising:

accessing, by the management system, a second plurality of historical component entries of the plurality of different entities, wherein each of the second plurality of historical component entries includes an identifier of a component of an object associated with the second historical component entry but does not include an identifier of a service performed on the object, the component of the object acquired by a respective user from a respective entity from the plurality of different entities after purchase of the object without the service being performed on the object at the respective entity; and

generating, by the management system, the training data by classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries into at least one of the plurality of predetermined component classifications based on the identifier of the component of the object included in each of the first plurality of historical component entries and the identifier of the component included in each second plurality of historical component entries.

16. The non-transitory computer-readable storage medium of claim 15 , wherein training the machine learning model comprises:

extracting features from each of the first plurality of historical component entries and each of the second plurality of historical component entries included in the training data, the extracted features including a predetermined component classification assigned to each of the first plurality of historical component entries and each the second plurality of historical component entries, features of the objects included in the first plurality of historical component entries, features of services performed on the objects included in the first plurality of historical component entries, and features of the components included the first plurality of historical component entries and the second plurality of historical component entries, wherein the machine learning model is trained using the extracted features.

17. The non-transitory computer-readable storage medium of claim 15 , wherein classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries comprises:

identifying, for each of the first plurality of historical component entries and each of the second plurality of historical component entries, a keyword included in the respective historical component entry;

comparing, for each of the first plurality of historical component entries and each of the second plurality of historical component entries, the keyword with the plurality of predetermined component classifications to identify a match between the keyword and at least one of the plurality of predetermined component classifications; and

classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries with their respective matching predetermined component classification.

18. The non-transitory computer-readable storage medium of claim 15 , wherein classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries comprises:

accessing a list of identifiers of components of the plurality of different entities, the list including a mapping of each of the identifiers of components to one or more of the plurality of predetermined component classifications;

comparing the identifier of the component included in each of the first plurality of historical component entries and each of the second plurality of historical component entries to the list, the comparison resulting in a match between identifiers of the components from the first plurality of historical component entries and the second plurality of historical component entries and the identifiers of components in the list; and

classifying each of the first plurality of historical component entries and each of the second plurality of historical component entries with the predetermined component classification based on the match.

19. The non-transitory computer-readable storage medium of claim 12 , wherein the management system provides the recommended set of components to a client device of the user a plurality of instances including responsive to the request for service, upon check-in of the user at the entity for the service, and user confirmation to begin servicing the object.

20. A system for generating recommended components for an object of a user, comprising:

one or more computer processors; and

a non-transitory computer-readable storage medium storing code, the code when executed by the one or more computer processors cause the one or more computer processors to perform steps comprising:

accessing, by a management system, a first plurality of historical component entries of a plurality of different entities, wherein each of the first plurality of historical component entries includes an identifier of a component of an object previously purchased by a respective user and an identifier of a service previously performed on the object at a respective entity from the plurality of different entities after the purchase of the object, the component of the object acquired in conjunction with the service being previously performed on the object;

training, by the management system, a machine learning model using training data that is based at least in part on the identifier of the component of the object and the identifier of the service included in each of the first plurality of historical component entries, the trained machine learning model configured to predict for each of a plurality of predetermined component classifications a likelihood of selection of a component corresponding to one of the plurality of predetermined component classifications;

receiving, by the management system, a request for a service to be performed on the object previously purchased by the user at an entity from the plurality of different entities, the request including attributes of the object that was previously purchased by the user and a specific service for the entity to perform on the object at the entity;

applying, by the management system, the attributes of the object to the trained machine learning model responsive to the request, the trained machine learning model outputting for each of the plurality of predetermined component classifications a prediction of the likelihood of user selection of the component corresponding to the predetermined component classification based on the attributes of the object applied to the trained machine learning model;

determining, by the management system, a recommended set of components for the object that are provided by the entity based on the prediction for each of the plurality of predetermined componenet classications; and

providing, by the management system, the recommended set of components for the object to acquire by the user in conjunction with the specific service that was requested by the user being performed on the object at the entity along with a response to the request for the service.