No-CAPTCHA Selective Information Reveal for Bot Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing CAPTCHA systems require humans to identify obfuscated characters, which can be difficult and lead to incorrect determinations, causing disruption in user experience and potential misclassification of humans as bots.

Innovation Solution

Implementing a CAPTCHA system that selectively reveals un-obfuscated information to users, allowing them to respond based on clear content while software agents struggle to distinguish it from other information, thereby improving the accuracy of human identification without requiring humans to decipher obfuscated characters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If obfuscated characters are used in CAPTCHA, then software agents have difficulty identifying them, but humans also have difficulty and may be misclassified as bots

Engineering Contradiction:

Improveaccuracy of human identificationVSAvoiduser experience

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The CAPTCHA challenge is segmented into multiple portions, where only certain portions contain the target information while others are distractors. This segmentation allows humans to focus on identifying the relevant portions without being overwhelmed by obfuscation, while software agents struggle to distinguish which portions contain meaningful information.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of making all characters obfuscated and asking users to identify them (traditional approach), the patent inverts the approach by presenting multiple portions where only some contain the answer. Users must identify which portions are relevant, reversing the traditional obfuscation paradigm and improving both accuracy and user experience.

Inventive Principle:

Principle #13The other way round (Inversion)

2Reliability

If obfuscated characters are used in CAPTCHA, then software agents are harder to distinguish from humans, but user experience is disrupted

Engineering Contradiction:

Improveaccuracy of bot detectionVSAvoiduser experience

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

Different portions of the CAPTCHA have different qualities - some portions contain clear, identifiable information while others are obfuscated or irrelevant. This local quality differentiation allows humans to easily identify the relevant portions, while software agents without the ability to discern quality differences struggle with the challenge.

Inventive Principle:

Principle #3Local quality

3Reliability

If traditional CAPTCHA methods are used, then bot detection is achieved, but errors occur in misclassifying humans as bots

Engineering Contradiction:

Improvebot detection accuracyVSAvoidhuman identification accuracy

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The system incorporates feedback mechanisms where the difficulty and composition of CAPTCHA challenges can be adjusted based on performance data. This allows the system to learn from misclassification errors and refine the distinction between human and bot responses, improving overall identification accuracy over time.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9465928B2No-CAPTCHA CAPTCHA

Publication Date: 2016.10.11 VERIZON PATENT & LICENSING INC

AI summary

A device generates parameters identifying selectively revealed un-obfuscated information associated with administering and assessing a Turing test. The device provides, to a client device, a challenge based on the parameters. The challenge directs the client device to selectively reveal the un-obfuscated information for presentation to a user associated with the client device. The device receives a response, to the challenge, from the client device and determines, based on the response and the parameters, whether the user associated with the client device is human. The device selectively performs an action based on determining whether the user associated with the client device is human.