System And Method For High Accuracy Product Classification With Limited Supervision

摘要

Systems and methods are disclosed herein for classifying records, such as product records, using a machine learning algorithm. After training a classification model according to a machine learning algorithm using an initial training set, records are classified and high confidence classifications identified. Remaining classifications are submitted to a crowdsourcing forum that validates or invalidates the classifications or marks them as to unclear to evaluate. Invalidated classifications are automatically analyzed to identify one or both of classification values and categories having a high proportion of invalidated classifications. Requests are transmitted to analysts to generate training data that is added to the training set. The process of classifying records and obtaining crowdsourced validation thereof may then repeat.

主权项

1. A method for classification, the method comprising:
training, by a computer system, a classification model using a training data set; classifying, by the computer system, using the classification model, a record set to generate a classification outcome set, the classification outcome set including classifier-record pairings for records of the record set; submitting, by the computer system, the record set and the classification outcome set to a crowdsourcing computer network; receiving, by the computer system, a validated portion and a non-validated portion of the classification outcome set from the crowdsourcing computer network; transmitting, by the computer system, the non-validated portion to an analyst computer network; receiving, by the computer system, from the analyst computer network, additional training data relating to classifiers of the classifier-record pairings; adding, by the computer system, the additional training data to the training data set.