Two-Tier Cognitive Verification for Bot Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current human verification methods, whether low-level or high-level cognitive tasks, are ineffective against advanced machine learning-based computer bots, as they either allow bots to pass easily or become too difficult for humans to navigate due to injected noise and complexities.
Innovation Solution
A two-tiered approach is implemented, where a first cognitive level question with fake images is used to differentiate humans from bots, and if the response confidence is below a threshold, a second, higher cognitive level question is provided, increasing complexity to enhance detection precision without sacrificing usability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a low-level cognitive task (text-based CAPTCHA) is used for verification, then the ease of operation is improved, but the reliability deteriorates because machine learning-based bots can solve it easily
Solution Approach 1:
The patent changes the parameter of cognitive complexity from low-level text recognition to high-level object recognition with spatial relationships. This parameter change makes the task difficult for machine learning bots while remaining solvable by humans, thus improving reliability without sacrificing ease of operation significantly
Solution Approach 2:
The verification system dynamically adjusts the type of cognitive task based on the detected capability of the requestor. By observing responses to initial questions, the system adapts the complexity and nature of subsequent questions, creating a dynamic verification process that optimizes both reliability and user experience
2Reliability
If a high-level cognitive task (object recognition) is used for verification, then the reliability is improved, but the ease of operation deteriorates due to increased complexity and noise
Solution Approach 1:
The verification process is segmented into multiple stages: initial low-level cognitive questions followed by progressive high-level cognitive questions. This segmentation allows the system to gradually assess capability without overwhelming the user immediately, improving ease of operation while maintaining reliability
Solution Approach 2:
The system applies partial action by not requiring all users to complete the full high-level cognitive assessment. Instead, it uses a tiered approach where only those who fail initial simpler questions proceed to more complex tasks, reducing the overall burden while maintaining detection accuracy
3Measurement precision
If machine learning techniques are used to improve detection accuracy, then the measurement precision is improved, but the device complexity increases due to noise injection and countermeasures
Solution Approach 1:
The patent introduces an intermediary cognitive assessment layer between the user and the final verification decision. This intermediary process uses structured questioning at multiple cognitive levels to mediate the detection process, simplifying the overall system complexity while maintaining high measurement precision through progressive capability assessment
Data Source
AI summary
An approach is provided in which the approach provides a first question to a requestor requesting access to a resource. The first question corresponds to a first cognitive level and includes at least one image selected from a set of images. The approach computes a confidence value of a first answer received from the requestor responding to the first question. In response to determining that the confidence value is below a confidence threshold, the approach provides a second question to the requestor corresponding to a second cognitive level that is increased from the first cognitive level. The approach grants access to the requestor in response to determining that a second answer received from the requestor responding to the second question is a correct answer.