Entropic Analysis for Credential Sharing Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately detect unauthorized credential usage due to the complexity of distinguishing between legitimate multi-device streaming and fraudulent activity, as they often penalize legal use cases involving multiple devices or locations.
Innovation Solution
The method involves determining watch-time variability by analyzing account and streaming data to generate a viewing probability distribution, grouping streams based on account-stream characteristics, and calculating group entropies to measure the increase in disorder, thereby indicating potential credential sharing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional credential sharing detection methods are used that monitor number of devices and streaming locations, then detection sensitivity to credential sharing increases, but false positive rate increases due to penalizing legitimate multi-device usage
Solution Approach 1:
The patent transforms the detection approach by changing from monitoring raw counts of devices and locations to calculating entropy-based metrics (watch-time variability) that capture the distribution and coordination patterns of viewing behavior. This parameter transformation allows differentiation between chaotic credential sharing and coordinated legitimate usage.
Solution Approach 2:
The patent replaces mechanical counting methods with information-theoretic entropy analysis. Instead of simply counting devices and locations, the system uses entropy calculations to measure the unpredictability and coordination of viewing patterns, substituting a more sophisticated analytical approach that captures behavioral nuances.
2Measurement precision
If monitoring of multiple devices and locations is implemented, then credential sharing detection capability improves, but system complexity increases
Solution Approach 1:
The patent extracts the essential characteristic of credential sharing behavior by isolating the entropy metric from the complex multi-dimensional data. Instead of analyzing all device and location data directly, the system extracts watch-time variability as a single discriminating feature that captures the essence of coordinated versus chaotic usage patterns.
3Measurement precision
If entropic analysis of streaming behavior is used, then differentiation between legitimate and unauthorized usage improves, but computational requirements increase
Solution Approach 1:
The patent applies partial action by calculating entropy metrics on sampled or aggregated streaming data rather than analyzing every individual stream event in real-time. This approach maintains high detection accuracy while reducing computational burden through selective analysis of representative data subsets.
Data Source
AI summary
A computer-implemented method for determining a credential sharing of a video streaming service account based on entropy of streaming data is described. The method includes determining an account entropy based on a viewing probability distribution for a total amount of content streamed for a defined account within a defined analysis period, grouping the total amount of content streamed into groups based on an account-stream characteristic that has probabilistic utility in determining the credential sharing, determining a group entropy for each of the groups, determining a watch-time variability based on the account entropy and each group entropy, and determining the credential sharing for the defined account based on the watch-time variability. The watch-time variability measures an increase in disorder when two or more groups of the groups are unrelated with respect to the account-stream characteristic.