Closed Captioning Classification for Ad and Program Segments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing content distribution systems struggle to accurately identify and classify segments of media content, such as advertisements and program segments, within a sequence of media content, which is crucial for understanding when and where advertisements are inserted, obtaining accurate copies of advertisements, and removing advertisements from media content for on-demand streaming.
Innovation Solution
A computing system retrieves closed captioning repetition data, generates features from this data, and uses a classification model, such as a neural network, to predict whether a segment is a program or advertisement segment, storing the prediction in a database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional content distribution systems are used to identify media content segments, then the system structure remains simple, but the classification accuracy of segments (advertisement vs. program) is insufficient
Solution Approach 1:
The patent introduces closed captioning data as an intermediary element to bridge the gap between simple content transmission and accurate segment classification. The closed captioning text serves as a mediator that contains linguistic information about the media content, enabling the classification model to distinguish between advertisement and program segments without requiring complex direct analysis of the media content itself
Solution Approach 2:
The patent replaces traditional mechanical/content-based analysis methods with an information-processing approach using closed captioning text and machine learning classification. Instead of analyzing video/audio content directly through complex signal processing, the system substitutes this with text-based feature extraction and classification model inference, achieving higher accuracy with comparable system complexity
2Measurement precision
If closed captioning data is used to classify media segments, then segment identification accuracy improves, but data processing complexity increases
Solution Approach 1:
The patent extracts only the necessary linguistic features from closed captioning data rather than processing the entire captioning dataset. The feature extraction process selectively identifies key characteristics (such as text patterns, repetition frequency, and linguistic structures) that are most indicative of advertisement versus program content, reducing processing complexity while maintaining high identification accuracy
Solution Approach 2:
The patent performs preliminary processing of closed captioning data by pre-extracting relevant features and preparing classification inputs before the actual segment classification occurs. This preliminary action organizes the raw captioning data into structured feature vectors that are optimized for the classification model, reducing the computational burden during real-time or near-real-time segment identification
Data Source
AI summary
In one aspect, an example method includes (i) retrieving, from a text index, closed captioning repetition data for a segment of a sequence of media content; (ii) generating features using the closed captioning repetition data; (iii) providing the features as input to a classification model, wherein the classification model is configured to output classification data indicative of a likelihood of the features being characteristic of a program segment; (iv) obtaining the classification data output by the classification model; (v) determining a prediction of whether the segment is a program segment using the classification data; and (vi) storing the prediction for the segment in a database.