ASR-Based High-Entropy Passphrases From Local-Language Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Basic Emergent Users (BEUs) face challenges in generating high entropy passphrases due to limited vocabulary and language barriers, making them vulnerable to dictionary attacks and social circle threats, with existing solutions failing to address their unique needs.
Innovation Solution
A method and system that utilizes Automated Speech Recognition (ASR) to convert spoken phrases into text, filters Seed List Words (SLWs) based on a personalized intersection corpus, generates a passphrase distance matrix, and displays high entropy passphrases at optimal screen positions for easy recall, ensuring robust device authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If BEUs use simple passphrases that are easy to recall, then ease of operation is improved, but security reliability deteriorates
Solution Approach 1:
The patent introduces an Automated Speech Recognition (ASR) engine as an intermediary that converts the user's spoken local language phrase into text and processes it through a personalized intersection corpus. This mediator bridges the gap between the user's limited vocabulary and the requirement for high-entropy passphrases, automatically generating secure passphrases from simple user input without requiring the user to understand entropy or vocabulary constraints.
Solution Approach 2:
The system transforms the passphrase generation process by changing parameters: it converts speech to text, filters through a personalized corpus, calculates vector distances between words, and generates multiple passphrase candidates with varying entropy levels. This parameter transformation allows the system to produce passphrases that are both memorable (based on user's own words) and secure (through automated entropy optimization).
2Reliability
If BEUs are provided with a large vocabulary for passphrase generation, then entropy is improved, but ease of operation deteriorates
Solution Approach 1:
The patent extracts only the necessary elements from the user's spoken phrase that are present in the personalized intersection corpus, filtering out unnecessary vocabulary. Instead of presenting the user with a large vocabulary, the system extracts and processes only the relevant words from the user's own speech, automatically handling the complexity of vocabulary selection and entropy calculation.
Solution Approach 2:
The system performs self-service by automatically generating multiple passphrase candidates with different entropy levels based on the user's spoken input. The ASR engine and corpus processing occur without user intervention, and the system autonomously calculates vector distances, ranks candidates by entropy, and presents options to the user, eliminating the need for the user to navigate complex vocabulary choices.
3Adaptability or versatility
If a personalized intersection corpus is created for each user, then adaptability is improved, but device complexity increases
Solution Approach 1:
The personalized intersection corpus is created in advance by collecting and processing the user's spoken phrases, storing the intersection of words between the user's vocabulary and a reference corpus. This preliminary action prepares the personalized data structure before passphrase generation is needed, so that during actual passphrase creation, the system only needs to perform vector distance calculations on pre-filtered data, reducing real-time processing complexity.
Data Source
AI summary
There are technical challenges to address the technical capability limitations of a Basic Emergent User (BEU) to enable and assist in generating high entropy passphrases but are easier for recall. Generation usable high entropy passphrases in local language from personalized intersection corpus for device authentication set up for BEU is provided. A recallable phrase, spoken by the BEU in local language, is converted to a text and Seed List Words (SLWs) are filtered based on a pre-generated personalized intersection corpus. Passphrase distance matrix is generated for the SLWs using the personalized intersection corpus. Words associated with each SLW are arranged in descending order of vector distance or entropy. Words of same order are concatenated in for each SLW to generate passphrase corpus. Randomly selected passphrases are read out and displayed on the user device by positioning the highest entropy based on usability of display screen in context of the user.