Media Program Similarity Computation via Metadata Vector Cross-Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in effectively correlating and quantifying similarities between media programs, particularly video and audio files, which are prevalent due to the complexity of non-textual data, limiting their ability to provide accurate recommendations and user experiences in media platforms like smart TV systems.
Innovation Solution
A computer-implemented method that extracts tokens from metadata fields of media programs, assigns weight factors based on frequency and occurrence, and calculates a similarity score through vector representation and dot products, allowing for cross-correlation and recommendation of similar media programs to users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to compare media programs, then the process is simple, but the similarity calculation is inaccurate for non-textual data
Solution Approach 1:
The patent introduces metadata as an intermediary layer between the original media programs and the comparison process. Instead of directly comparing complex non-textual media content, the system extracts and compares structured metadata (tokens, fields, vectors) that represent essential characteristics of the media programs. This intermediary representation enables accurate similarity calculation while maintaining manageable processing complexity.
Solution Approach 2:
The patent replaces traditional mechanical comparison methods with mathematical vector operations. Media programs are transformed into vector representations where similarity can be calculated using dot products and other mathematical operations. This substitution enables precise measurement of similarity between non-textual data by converting it into a mathematical framework that supports accurate computation.
2Measurement precision
If detailed metadata analysis is performed, then similarity accuracy improves, but processing time increases
Solution Approach 1:
The patent segments the metadata analysis process into distinct fields (e.g., cast, description, title) and extracts tokens from each field separately. This segmentation allows the system to process and compare specific aspects of media programs independently, enabling selective analysis that balances accuracy with processing efficiency. Not all fields need to be analyzed in full detail for every comparison.
Solution Approach 2:
The patent transforms metadata into vector representations with adjustable parameters. By changing the parameter representation (from raw text to weighted vectors), the system enables efficient mathematical operations that reduce processing time while maintaining measurement accuracy. The vector format allows for rapid computation of similarity metrics compared to traditional text analysis methods.
Data Source
AI summary
System and method are provided to associate or compare media programs. A method includes: obtaining, using at least one processing circuit, first metadata for a first media program and second metadata for a second media program, wherein the first metadata are organized into a plurality of first fields, and the second metadata are organized into a plurality of second fields; extracting, using at least one processing circuit, a plurality of first tokens from one of the plurality of the first fields and a plurality of second tokens from one of the plurality of second fields; assigning a weight factor to each of the first and second tokens; cross-correlating the first and second tokens between the plurality of first fields and the plurality of second fields; and calculating a similarity score between the first and second media programs based on the cross-correlating.