Account Sharing Detection via Risk Scoring and Location Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Streaming service providers face challenges in identifying and addressing impermissible account sharing, as existing methods are inefficient and often incorrectly flag legitimate users, leading to revenue loss and increased costs due to unstructured and noisy data.
Innovation Solution
A method and system for detecting impermissible account sharing by determining locations accessed by a user account, calculating a risk coefficient for each location based on device usage and distance from a base location, and generating a sharing score to flag accounts exceeding a threshold, which efficiently processes large data volumes and provides explainable results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional data processing methods are used to identify shared accounts, then comprehensive analysis can be performed, but processing time becomes extremely long and computational resources are excessive
Solution Approach 1:
The patent segments the account sharing detection process into multiple independent components: device fingerprinting module, location tracking module, usage pattern analysis module, and risk scoring module. Each module processes specific aspects of account behavior independently, allowing parallel processing and significantly reducing overall computation time while maintaining comprehensive analysis capability
Solution Approach 2:
The system performs preliminary actions by continuously collecting and preprocessing user data in the background before actual sharing detection is needed. Device fingerprints are generated, locations are tracked, and usage patterns are established as baseline data, so when account sharing detection is triggered, the system can quickly compare against pre-prepared data rather than processing raw data from scratch
2Measurement precision
If comprehensive user data is collected for accurate sharing detection, then detection accuracy improves, but data complexity and noise increase making parsing difficult
Solution Approach 1:
The patent extracts only the most relevant features from comprehensive user data for sharing detection. Instead of processing all raw user data, the system extracts specific indicators such as device fingerprint characteristics, location coordinates, usage time patterns, and content access patterns. This extraction process filters out noise and irrelevant information while preserving the essential signals needed for accurate detection
Solution Approach 2:
The system applies different processing qualities to different data elements based on their relevance. High-precision processing is applied to critical indicators like device fingerprints and location data, while less critical data elements undergo lighter processing. This localized quality approach optimizes computational resources and reduces overall data complexity while maintaining detection accuracy
3Loss of energy
If account sharing is strictly prohibited, then revenue protection is maximized, but legitimate household sharing is incorrectly flagged causing user dissatisfaction
Solution Approach 1:
The patent implements dynamic detection thresholds and risk scoring that adapt based on multiple factors. Rather than applying a static rule that flags all multi-device usage as sharing, the system dynamically adjusts detection sensitivity based on usage patterns, time of day, geographic proximity, and account history. This dynamic approach allows legitimate household sharing to pass through while capturing actual impermissible sharing
Solution Approach 2:
The system incorporates feedback mechanisms where detection results are continuously refined based on user responses and additional data. When accounts are flagged, the system can request additional verification information or adjust its scoring based on user feedback about legitimate sharing scenarios. This feedback loop reduces false positives over time while maintaining revenue protection
Data Source
AI summary
The present disclosure relates to a method for detecting impermissible account sharing among user accounts of a streaming media service including the steps of determining a plurality of locations accessed by a given user account of the user accounts; determining a device access count for each of the locations, the device access count indicating how many times the corresponding location was accessed by at least one device associated with the given user account; identifying one of the locations having the highest device access count as a base location; calculating a risk coefficient for each remaining location; generating a sharing score for the given user account by summing the risk coefficients; and determining impermissible account sharing of the given user account has occurred when the sharing score exceeds a threshold.