Stationary User Pair Identification via Location Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network service providers face challenges in identifying users who know each other based solely on location information, as proximity does not necessarily imply familiarity, and existing methods rely on direct evidence of communication which may be restricted by privacy rules.
Innovation Solution
A method using a statistical approach to generate a score for each user pair based on temporal, distance, and location parameters, determining the likelihood of users knowing each other without relying on direct evidence, and filtering out location data from users deemed to be moving, to identify stationary users who are likely to know each other.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If location information of all users is analyzed to identify pairs who know each other, then the completeness of user pair identification is improved, but the computational complexity and data processing burden increase significantly
Solution Approach 1:
The system segments users into two distinct groups: stationary users and moving users. This segmentation is based on analyzing location information over time to determine whether users remain at fixed locations or are in transit. By dividing the user population into these categories, the system reduces computational complexity while maintaining identification accuracy, as the patent demonstrates that pairs of stationary users have higher likelihood of knowing each other compared to moving users
Solution Approach 2:
The system extracts and isolates location information specifically from stationary users, separating this subset from the complete user dataset. This extraction process removes irrelevant data from moving users who are less likely to have meaningful interactions, thereby reducing the data processing burden while preserving the key information needed for accurate user pair identification
2Quantity of substance
If location data from all users including moving users is processed, then the coverage of potential user pairs is improved, but the quality of identification results deteriorates due to false positives
Solution Approach 1:
The system applies different quality criteria to different user segments. Specifically, it recognizes that stationary users represent a higher quality segment for identification purposes, as they are more likely to have meaningful interactions. The system selectively focuses analysis on this high-quality segment rather than treating all users uniformly, thereby improving identification accuracy without sacrificing meaningful coverage
3Measurement precision
If direct evidence of communication is used to identify users who know each other, then the accuracy of identification is improved, but privacy restrictions and data availability limitations worsen
Solution Approach 1:
Instead of directly analyzing communication evidence to identify users who know each other, the system inverts the approach by analyzing the absence of movement (stationary behavior) as a proxy indicator. Rather than seeking direct proof of interaction, the system uses the inverse logic that users who remain at fixed locations together are likely to know each other, thereby achieving identification accuracy without requiring direct communication data
Data Source
AI summary
A method and apparatus for identifying a plurality of users of a service of a network service provider as users known to each other are disclosed. For example, the method implemented via a processor obtains location information of a plurality of users of a service of a network service provider, selects a subset of users from the plurality of users that are deemed to be stationary, generates a score for each particular pair of users selected from the subset of users, determines for each particular pair of users, whether the score that is generated is greater than or equal to a threshold, identifies the pair of users as users who know each other, when the score for the particular pair of users is greater than or equal to the threshold, and performs an analysis to provide an offer to the pair of users, wherein the offer is based on the pair of users being identified as users who know each other.