Voiceprint Recognition for Shared Account User Identification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing user recommendation systems on platforms are not accurate when multiple users share the same account, as they aggregate historical data without distinguishing individual user preferences, leading to inappropriate service recommendations.

Innovation Solution

An interaction method that uses audio recording and speech recognition to determine user identity through unique voice characteristics, allowing for personalized service recommendations by generating speech feature vectors and matching them with user profiles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple users share the same account information, then the platform can serve multiple users through a single account, but the recommendation accuracy deteriorates because aggregated historical data cannot distinguish individual user preferences

Engineering Contradiction:

Improveaccount sharing capabilityVSAvoidrecommendation accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments user identification from account information by introducing voiceprint recognition. Instead of relying solely on account data, the system divides the identification process into account verification and voiceprint matching, allowing separate handling of shared account scenarios and individual user preferences.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces voiceprint information as an intermediary element between the user and the recommendation system. This intermediary enables the system to distinguish individual users within shared accounts by matching voice characteristics against stored voiceprints, thereby resolving the accuracy issue without limiting account sharing.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If the system collects and analyzes more user data to improve recommendation accuracy, then service precision improves, but user privacy protection becomes more challenging

Engineering Contradiction:

Improveservice precisionVSAvoidprivacy risk

Core Design Contradiction:

Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent extracts voiceprint information from general user data collection. By specifically collecting only voice characteristics needed for identification rather than comprehensive user behavior data, the system achieves accurate user distinction while minimizing privacy intrusion and data security risks.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the identification parameter from broad user behavior data to specific voiceprint biometric data. This parameter shift enables precise user identification with smaller, more targeted data collection, reducing the privacy and security concerns associated with extensive data gathering.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11238870B2Interaction method, electronic device, and server

Publication Date: 2022.02.01 ALIBABA GROUP HOLDING LTD

AI summary

The present application discloses a method, device, system, and computer system for interfacing with a terminal in connection with performing a service function. The method includes receiving audio information based at least in part on a voice input from a user, determining a service function corresponding to the audio information, determining user information for the user based at least in part on the audio information, and performing the service function based at least in part on the user information.