US 11,769,165 B2Granted

Integrated explicit intent and inference based job seeker identification and segmentation

AssigneeMicrosoft Technology Licensing, LLC

InventorsChunzhe Zhang, Satej Milind Wagle, Linda Fayad, Ada Cheuk Ying Yu

Life of this patent

Estimating the term…

Loading patent…

Abstract

In an example embodiment, a specialized machine learned model, called a look-alike model, is trained using a machine learning algorithm to predict future job engagement for a user. This look-alike model is then used to create new segments on top of the segments provided by a rules-based approach. Specifically, the look-alike model is designed to take users who have been segmented by a rule-based approach into an “inactive job seeker” categorization (such as those assigned to the resting users and dormant users segments) and calculate a predicted job engagement score for these users. Based on the predicted job engagement score, a user may then be reassigned from one of the inactive job seeker categorizations to one of one or more new job seeker categorizations (such as predicted open job seekers or predicted opportunistic job seekers).

Claims19

Dependent claims

A system for training and using a machine learned model, comprising:

a processor;
a memory storing instructions, when executed by the processor, implement:
a training data preparation component configured to obtain training data, the training data including data from one or more user profiles in an online network and corresponding job engagement data, the job engagement data including information about interactions, via a graphical user interface of the online network, between corresponding users and job-related content in the online network, and to transform the training data;
a machine learning algorithm configured to train a look-alike machine learned model using the transformed training data, to cause the look-alike machine learned model to take information about a user and calculate a job engagement score for the user, the job engagement score being a prediction of future interactions, and via the graphical user interface of the online network, between the user and job-related content in the online network, within a predetermined time period;
an initial segmentation component configured to obtain information about a first user, assign the first user to an initial segment of users based on the information about the first user, the initial segment being one of a plurality of segments of users, determine whether the initial segment of users matches a segment of users defined as a variable segment, and, in response to a determination that the initial segment matches the segment of users defined as a variable segment, cause the look-alike machine learned model to calculate a job engagement score for the first user by passing the information about the first user to the look-alike machine learned model;
a final segmentation component configured to, prior to presenting content to the first user based on the assignment of the first user to an initial segment, re-assign the first user to a final segment of users different than the initial segment of users, based on the calculated job engagement score for the first user;
a retraining component configured to compare actual job engagement of the first user over the predetermined time period with the final segment of users to which the first user was assigned, and retrain the look-alike machine learned model based on the comparison.

4 more elements

A computerized method comprising:

obtaining, by a processor, training data including data from one or more user profiles in an online network and corresponding job engagement data, the job engagement data including information about interactions, via a graphical user interface of the online network, between corresponding users and job-related content in the online network;
transforming the training data;
training a look-alike machine learned model, with a machine learning algorithm, using the transformed training data, to cause the look-alike machine learned model to take information about a user and calculate a job engagement score for the user, the job engagement score being indicative of a prediction of future interactions, via the graphical user interface of the online network, between the user and job-related content in the online network, within a predetermined time period;
obtaining information about a first user;
assigning the first user to an initial segment of users based on the information about the first user, the initial segment being one of a plurality of segments of users;
determining whether the initial segment of users matches a segment of users defined as a variable segment;
in response to a determination that the initial segment matches the segment of users defined as a variable segment, calculating a job engagement score for the first user by passing the information about the first user to the look-alike machine learned model;
prior to presenting content to the first user based on the assignment of the first user to an initial segment, re-assigning the first user to a final segment of users different than the initial segment of users, based on the calculated job engagement score for the first user;
comparing actual job engagement of the first user over the predetermined time period with the final segment of users to which the first user was assigned;
and
retraining the look-alike machine learned model based on the comparison.

8 more elements

A non-transitory machine-readable storage medium comprising instructions which, when implemented by one or more machines, cause the one or more machines to perform operations comprising:

obtaining training data, the training data including data from one or more user profiles in an online network and corresponding job engagement data, the job engagement data including information about interactions, via a graphical user interface of the online network, between corresponding users and job-related content in the online network;
transforming the training data;
training a look-alike machine learned model, with a machine learning algorithm, using the transformed training data, to cause the look-alike machine learned model to take information about a user and calculate a job engagement score for the user, the job engagement score being indicative of a prediction of future interactions, via the graphical user interface of the online network, between the user and job-related content in the online network, within a predetermined time period;
obtaining information about a first user;
assigning the first user to an initial segment of users based on the information about the first user, the initial segment being one of a plurality of segments of users;
determining whether the initial segment of users matches segment of users defined as a variable segment;
in response to a determination that the initial segment matches the segment of users defined as a variable segment, calculating a job engagement score for the first user by passing the information about the first user to the look-alike machine learned model;
prior to presenting content to the first user based on the assignment of the first user to an initial segment, re-assigning the first user to a final segment of users different than the initial segment of users, based on the calculated job engagement score for the first user;
comparing actual job engagement of the first user over the predetermined time period with the final segment of users to which the first user was assigned;
and
retraining the look-alike machine learned model based on the comparison.

8 more elements

Dependencies are inferred from the claim text.