Logistic recommendation engine

1. A method comprising:

receiving, for a package and from a graphical user interface, shipment details comprising a plurality of attributes;

obtaining, for a subset of the plurality of attributes, a plurality of logistic preferences;

applying the plurality of logistic preferences to the shipment details to obtain modified shipment details;

training a classifier, of a plurality of classifiers, using a plurality of shipment transactions comprising a shipment transaction comprising a plurality of values for the plurality of attributes and labeled with a vendor logistic service of a plurality of vendor logistic services;

generating, by applying the classifier to the modified shipment details, a plurality of scores for the plurality of vendor logistic services,

wherein the classifier is one of the plurality of classifiers that are applied to the modified shipment details, and

wherein a score of the plurality of scores is a weighted combination of a plurality of classifier outputs identifying probabilities of matches, between the plurality of vendor logistic services and the modified shipment details, and generated using a harmonic mean in the modified shipment details; and

presenting a recommendation of the vendor logistic service, from the plurality of vendor logistic services, using the score generated from the output, from a classifier of the plurality of classifiers, that identifies a probability that the vendor logistic service matches to the modified shipment details.

2. The method of claim 1 , wherein the plurality of attributes comprises an origin location of the package and a destination location of the package.

3. The method of claim 1 , further comprising:

obtaining, by applying a scanning device to the package, a value for an attribute of the plurality of attributes, wherein the attribute is external to the subset of the plurality of attributes.

4. The method of claim 1 , wherein the plurality of logistic preferences comprises preferences corresponding to the package, preferences corresponding to a sender of the package, and preferences corresponding to a recipient of the package.

5. The method of claim 1 , further comprising:

obtaining a plurality of shipment transactions each comprising a value for an attribute of the plurality of attributes, wherein the attribute corresponds to a logistic preference of the plurality of logistic preferences; and

generating the logistic preference by calculating the harmonic mean of the values for the attribute in the plurality of shipment transactions.

6. The method of claim 1 , further comprising:

training the plurality of classifiers using different subsets of the plurality of shipment transactions, wherein generating the plurality of scores comprises:

applying the plurality of classifiers to the modified shipment details to obtain the plurality of classifier outputs; and

combining, according to a policy, the plurality of classifier outputs.

7. The method of claim 1 , further comprising:

displaying, in an element within a graphical user interface (GUI) generated by a computer processor, the recommendation of the vendor logistic service; and

initiating, via the GUI, an interaction with the vendor logistic service.

8. A system, comprising:

a computer processor;

a repository configured to store:

shipment details comprising a plurality of attributes,

a plurality of logistic preferences for a subset of the plurality of attributes, and

a plurality of shipment transactions comprising a shipment transaction comprising a plurality of values for the plurality of attributes and labeled with a vendor logistic service of a plurality of vendor logistic services; and

a recommendation engine, executing on the computer processor, configured to:

receive, for a package and from a graphical user interface, the shipment details, obtain the plurality of logistic preferences for the subset of the plurality of attributes, apply the plurality of logistic preferences to the shipment details to obtain modified shipment details,

train a classifier, of a plurality of classifiers, using the plurality of shipment transactions,

generate, by applying the classifier to the modified shipment details, a plurality of scores for the plurality of vendor logistic services,

wherein the classifier is one of the plurality of classifiers that are applied to the modified shipment details, and

wherein a score of the plurality of scores is a weighted combination of a plurality of classifier outputs identifying probabilities of matches, between the plurality of vendor logistic services and the modified shipment details, and generated using a harmonic mean in the modified shipment details; and

present a recommendation of the vendor logistic service from the plurality of vendor logistic services using the score generated from the output, from a classifier of the plurality of classifiers, that identifies a probability that the vendor logistic service matches to the modified shipment details.

9. The system of claim 8 , wherein the plurality of attributes comprises an origin location of the package and a destination location of the package.

10. The system of claim 8 , wherein the recommendation engine is further configured to:

obtain, by applying a scanning device to the package, a value for an attribute of the plurality of attributes, wherein the attribute is external to the subset of the plurality of attributes.

11. The system of claim 8 , wherein the plurality of logistic preferences comprises preferences corresponding to the package, preferences corresponding to a sender of the package, and preferences corresponding to a recipient of the package.

12. The system of claim 8 , wherein the recommendation engine is further configured to:

obtain a plurality of shipment transactions each comprising a value for an attribute of the plurality of attributes, wherein the attribute corresponds to a logistic preference of the plurality of logistic preferences, and

generate the logistic preference by calculating the harmonic mean of the values for the attribute in the plurality of shipment transactions.

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

train the plurality of classifiers using different subsets of the plurality of shipment transactions, wherein generating the plurality of scores comprises:

applying the plurality of classifiers to the modified shipment details to obtain the plurality of classifier outputs; and

combining, according to a policy, the plurality of classifier outputs.

14. The system of claim 8 , further comprising a graphical user interface (GUI), wherein the recommendation engine is further configured to:

display, in an element within the GUI generated by the computer processor, the recommendation of the vendor logistic service; and

initiate, via the GUI, an interaction with the vendor logistic service.

15. A method comprising:

receiving, for a package and via a graphical user interface (GUI) generated by a computer processor, shipment details comprising a plurality of attributes;

obtaining, for a subset of the plurality of attributes, a plurality of logistic preferences;

sending the shipment details and the plurality of logistic preferences to a recommendation engine, wherein the recommendation engine:

applies the plurality of logistic preferences to the shipment details to obtain modified shipment details;

trains a classifier, of a plurality of classifiers, using a plurality of shipment transactions comprising a shipment transaction comprising a plurality of values for the plurality of attributes and labeled with a vendor logistic service of a plurality of vendor logistic services;

generates, by applying the classifier to the modified shipment details, a plurality of scores for the plurality of vendor logistic services,

wherein the classifier is one of the plurality of classifiers that are applied to the modified shipment details, and

wherein a score of the plurality of scores is a weighted combination of a plurality of classifier outputs identifying probabilities of matches, between the plurality of vendor logistic services and the modified shipment details, and generated using a harmonic mean in the modified shipment details; and

presents a recommendation of the vendor logistic service from the plurality of vendor logistic services using the score generated from the output, from a classifier of the plurality of classifiers, that identifies a probability that the vendor logistic service matches to the modified shipment details;

receiving, via the GUI, the recommendation of the vendor logistic service;

displaying, in an element within the GUI, the recommendation of the vendor logistic service; and

initiating, via the GUI, an interaction with the vendor logistic service.

16. The method of claim 15 , further comprising:

determining that a recommendation has been requested for the modified shipment details, wherein the modified shipment details are sent to the recommendation engine in response to the determining.

17. The method of claim 15 , wherein the plurality of attributes comprises an origin location of the package and a destination location of the package.

18. The method of claim 15 , further comprising:

obtaining, by applying a scanning device to the package, a value for an attribute of the plurality of attributes, wherein the attribute is external to the subset of the plurality of attributes.

19. The method of claim 15 , wherein the plurality of logistic preferences comprises preferences corresponding to the package, preferences corresponding to a sender of the package, and preferences corresponding to a recipient of the package.

20. The method of claim 15 , further comprising:

obtaining a plurality of shipment transactions each comprising a value for an attribute of the plurality of attributes, wherein the attribute corresponds to a logistic preference of the plurality of logistic preferences; and

generating the logistic preference by calculating the harmonic mean of the values for the attribute in the plurality of shipment transactions.