CONTENT RATING CLASSIFICATION WITH COGNITIVE COMPUTING SUPPORT

• US 20180276565A1
• Filed: 12/31/2017
• Published: 09/27/2018
• Est. Priority Date: 03/21/2017
• Status: Active Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Brandao, Rafael Rossi de Mello, Cerqueira, Renato Fontoura de Gusmao, Moreno, Marcio Ferreira
• Application Number:
  US15/859,578
• Publication Number:
  US 20180276565A1
• Time in Patent Office: Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/025 Extracting rules from data

Abstract

A method for classifying content includes receiving the content; identifying a ratings jurisdiction and regime for the content; accessing a knowledge base for a trained model according to the ratings jurisdiction and regime; classifying the content by testing the content against the trained model; and providing to a user the classification of the content. Optionally, the method includes receiving, from the user, a classification feedback of the content; and updating the trained model responsive to the content and the classification feedback.

Claims

1. A method comprising:

receiving content;

identifying a ratings jurisdiction and regime for the content;

accessing a knowledge base for a trained model according to the ratings jurisdiction and regime;

classifying the content by testing the content against the trained model; and

providing, to a user, the classification of the content.

2. The method of claim 1 further comprising:

receiving, from the user, a classification feedback of the content; and

updating the trained model responsive to the content and the classification feedback.

3. The method of claim 2 wherein the knowledge base includes content samples and updating the trained model includes adding at least portions of the content to the knowledge base as training data for the trained model.

4. The method of claim 1 wherein the content is bundled content.

5. The method of claim 4 further comprising decomposing the bundled content to obtain items of content, classifying each of the items of content, and aggregating the classifications of the items of content.

6. The method of claim 1 wherein the trained model includes a collection of classification rules that are trained on content samples and classification ratings of the content samples.

7. The method of claim 6 wherein the classification rules are trained on extracted semantics of the content samples, and testing the content against the trained model includes applying the classification rules to extracted semantics of the content.

8. The method of claim 6 wherein the classification rules are trained on feature vectors of the content samples, and testing the content against the trained model comprises applying the classification rules to feature vectors of the content.

9. The method of claim 1 wherein the knowledge base includes content samples.

10. The method of claim 9 further comprising extracting feature vectors from the content and obtaining ratings feature vectors of the content samples, wherein the trained model is trained on the ratings feature vectors of the content samples, and testing the content against the trained model comprises applying the trained model to the feature vectors of the content.