Anonymous User Identification via Behavioral Feature Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In online communities where anonymity is guaranteed, identifying and tracking the same user across nickname changes is challenging due to the lack of personal information and continuity, hindering data analysis and potential abuse of anonymity.
Innovation Solution
A method and apparatus that extract nicknames of identical users by analyzing feature information such as word-related, activity time, posting-related, and communication relationship features, using morpheme analysis and genetic algorithms to determine similarity and weight values, allowing for the identification of users even after nickname changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If anonymity is guaranteed in online community, then users can post frank opinions, but personal information cannot be identified
Solution Approach 1:
The patent introduces an intermediary identification system that uses behavioral features (posting patterns, communication relationships, activity timing) as mediators to link anonymous nicknames to user identities without exposing personal information directly. This allows opinion freedom to be maintained while enabling data identification through indirect behavioral markers.
Solution Approach 2:
The system changes the identification parameters from direct personal information to behavioral feature parameters such as posting frequency, communication patterns, and activity timing. By transforming the identification basis from static personal data to dynamic behavioral parameters, both anonymity and identifiability are simultaneously achieved.
2Adaptability or versatility
If nickname changes are allowed, then anonymity is maintained, but continuity with past identity is not guaranteed
Solution Approach 1:
The system employs feedback mechanisms where behavioral features are continuously tracked across nickname changes. By monitoring consistent behavioral patterns (communication relationships, posting habits, activity timing) over time, the system provides feedback that maintains identity continuity even as nicknames change, allowing both flexibility and reliability.
Solution Approach 2:
The patent establishes preliminary behavioral feature profiles when users first join, which serve as reference points for future identification. These pre-established behavioral markers enable continuous tracking across nickname changes by comparing future behavior against the preliminary profile, maintaining identity continuity through proactive pattern recognition.
3Measurement precision
If feature information analysis is performed to identify users, then user identification accuracy improves, but system complexity increases
Solution Approach 1:
The patent segments the identification system into distinct modular components: feature extraction modules (posting features, communication features, timing features), similarity calculation modules, and matching modules. This segmentation allows each component to handle specific tasks independently, improving identification accuracy through specialized processing while managing complexity through modular architecture.
Solution Approach 2:
The system transforms complex behavioral data into simplified feature parameters (posting frequency, communication patterns, activity timing). By changing the parameter representation from raw complex data to standardized feature metrics, identification accuracy is improved through consistent parameter comparison while system complexity is reduced through parameter normalization and standardization.
Data Source
AI summary
A method of extracting nicknames of identical user by an apparatus operated by at least one processor, the method includes receiving a posting uploaded to an online community from a server; extracting at least one feature information for identifying a posting writer who writes the posting, from the posting; and extracting nicknames of identical user with the posing writer, from a plurality of nicknames, based on similarity of the feature information with a predetermined reference or greater.