Non-Human User Identification via Multi-Dimensional Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need for effective techniques to identify illegitimate non-human user software accessing content, as existing methods struggle to distinguish between human and non-human traffic, particularly in advertising models like CPM, CPC, and CPA.
Innovation Solution
A method involving the selection of attributes relevant to non-human user software activity from a plurality of computerized devices, computing scores for these attributes, and combining them to determine a likelihood of non-human user software infection, using data from infected and non-infected devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to identify non-human users, then the system is simple to operate, but the identification accuracy is low and cannot effectively distinguish between human and non-human traffic
Solution Approach 1:
The patent segments the identification process into multiple independent scoring factors (device attributes, browsing behavior, interaction patterns, temporal patterns) that can be evaluated separately and then combined. This allows the complex identification task to be broken down into manageable components while maintaining high accuracy through multi-dimensional analysis.
Solution Approach 2:
The patent introduces multiple new dimensions for analysis beyond simple device identification, including browsing behavior patterns, interaction patterns with content, temporal patterns of activity, and device attribute correlations. This multi-dimensional approach significantly improves identification accuracy by examining traffic from multiple angles simultaneously.
2Measurement precision
If multi-dimensional analysis is performed to improve identification accuracy, then the measurement precision improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent performs preliminary actions by pre-establishing correlations between device attributes and non-human user software, and by defining scoring thresholds and weightings in advance. This allows the system to quickly evaluate new traffic by comparing it against pre-computed criteria rather than performing full analysis from scratch, significantly reducing processing time while maintaining accuracy.
Solution Approach 2:
The patent transforms qualitative behavioral observations into quantitative scored parameters with specific thresholds. By converting complex behavioral patterns into numerical scores that can be rapidly compared against predetermined thresholds, the system achieves fast processing while maintaining high identification accuracy through mathematical evaluation.
3Reliability
If comprehensive attributes are analyzed to identify non-human users, then the identification reliability improves, but the amount of data processing and system resources required increase
Solution Approach 1:
The patent extracts and focuses only on the most discriminative attributes and behavioral patterns that are most strongly correlated with non-human user software. By selecting and analyzing only the most relevant features rather than processing all possible data points, the system achieves high reliability while minimizing computational resource consumption.
Solution Approach 2:
The patent applies different levels of analysis depth to different attributes based on their discriminative power. High-value attributes that strongly indicate non-human behavior receive more detailed analysis, while less informative attributes receive lighter processing. This localized quality approach optimizes resource allocation while maintaining overall identification reliability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Improved techniques can be used to identify illegitimate non-human user software that is accessing content. An exemplary method of identifying non-human user software of computerized devices may comprise receiving information relating to attributes relevant to the indication of non-human user software activity from a plurality of computerized devices, wherein at least a portion of the computerized devices are known to be infected with at least one non-human user software, and at least a portion of the devices are known not to be infected with a non-human user software, selection as factors of the attributes based on a correlation of the attribute with the presence of non-human user software activity, computing a score for each factor indicating a likelihood of non-human user software infection for that factor, computing a combined score based on the scores of the individual factors, the combined score indicating a combined likelihood of non-human user software infection.