Enhanced data labeling for machine learning training

A method comprising:

• sending (i), to a first user device, a first set of target objects to be labeled and a first context set selected from a plurality of context sets related to the first set of target objects, and (ii), to a second user device, the first set of target objects and a second context set of the plurality of context sets, wherein the plurality of context sets comprise context objects that are related to the first set of target objects and are selected based on a use case of a machine learning model to be trained;
• storing, in a database, for training of the machine learning model, a first training dataset comprising first labels assigned to the first set of target objects based in connection with the first context set and a second training dataset comprising second labels assigned to the first set of target objects in connection with the second context set;
• executing, on a test dataset, a first instance of the machine learning model, trained on the first training dataset, to obtain a first accuracy score and a second instance of the machine learning model trained, on the second training dataset, to obtain a second accuracy score;
• based on the first accuracy score being greater than the second accuracy score, propagating one or more characteristics of the first context set to generate a third context set for labeling second target objects;
• providing, to the first user device and the second user device in connection with a labeling task for the second target objects the third context set, to obtain a third training dataset comprising the second target objects and a set of labels for the second target objects;
• and
• training a third instance of the machine learning model using the third training dataset.

One or more non-transitory computer-readable media comprising instructions that, when executed by one or more processors, cause operations comprising:

• storing, in a database for training a machine learning model, (i) a first training dataset comprising one or more target objects and first labels that are assigned to the one or more target objects based on a first set of context objects related to the one or more target objects and that are selected based on a use case of the machine learning model (ii) a second training dataset comprising the one or more target objects and second labels that are assigned to the one or more target objects based on a second set of context objects related to the one or more target objects and that are selected based on the use case of the machine learning model;
• training a first instance of the machine learning model on the first training dataset and a second instance of the machine learning model on the second training dataset to obtain a first trained machine learning model and a second trained machine learning model, respectively;
• executing the first trained machine learning model and the second trained machine learning model on a testing dataset to calculate a first performance score and a second performance score, respectively;
• based on the first performance score being greater than the second performance score, generating a third set of context objects for labeling a set of target objects using characteristics of the first set of context objects;
• and
• training a third instance of the machine learning model using a third training dataset comprising a set of target objects and third labels that are assigned to the set of target objects based on a third set of context objects derived using characteristics of the first set of context objects.