Caller Profile Generation for Shared-Number Caller Disambiguation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing crisis call centers struggle to identify repeat callers without compromising confidentiality, as traditional methods like phone numbers are shared among multiple users, leading to delayed responses and inefficient use of network and computing resources.

Innovation Solution

Implement a system with a caller profile database and a call analyzer that uses audio and text analytics to generate and identify caller profiles, associating them with existing profiles based on call characteristics, while maintaining anonymity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If phone numbers are used to identify callers, then caller identification is simple, but multiple users share the same number leading to inability to distinguish repeat callers

Engineering Contradiction:

Improvecaller identificationVSAvoidcaller distinction accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent segments the caller identification process into multiple independent analysis channels: audio fingerprinting of voice characteristics, text analytics of call content, and metadata analysis. Each channel extracts distinct features that collectively create a comprehensive caller profile, allowing differentiation of users sharing the same phone number.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system transforms the identification basis from a single parameter (phone number) to multiple parameters including audio spectral features, text linguistic features, call timing patterns, and interaction behaviors. This multi-parameter approach enables precise caller distinction while maintaining system simplicity.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If audio and text analytics are implemented to identify repeat callers, then caller identification accuracy improves, but system complexity and computing resource requirements increase

Engineering Contradiction:

Improverepeat caller identification accuracyVSAvoidsystem architecture

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary analysis by extracting and storing audio fingerprints and text features during the initial call interaction. These pre-computed features are saved in caller profiles for rapid comparison during subsequent calls, avoiding the need for complex real-time analysis and reducing overall system complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates simplified copies of caller characteristics through audio fingerprinting and text feature extraction. These digital representations serve as lightweight identifiers that can be rapidly compared without requiring full audio or text processing, reducing computational burden while maintaining identification accuracy.

Inventive Principle:

Principle #26Copying

3Productivity

If comprehensive caller profiling is performed, then resource optimization improves, but processing time and computational load increase

Engineering Contradiction:

Improvecall handling efficiencyVSAvoidcall processing time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system implements a tiered profiling approach where essential identification features (audio fingerprint, basic text features) are processed for all calls, while more comprehensive analysis is performed only when needed for repeat caller confirmation or complex situations. This partial action approach optimizes the balance between profiling completeness and processing speed.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12395585B2Disambiguation of multiple callers using a shared communication source

Publication Date: 2025.08.19 MOTOROLA SOLUTIONS INC

AI summary

Systems and methods for automated caller profile generation and identification. One example method includes receiving, from a service answering point, a service call from a caller. The method includes determining, with an electronic processor, based on a content of the service call, a call characteristic. The method includes generate a preliminary caller profile based on the call characteristic. The method includes comparing the preliminary caller profile to each of a plurality of existing caller profiles retrieved from a database storing caller profile data for a plurality of callers to generate, for each of the plurality of existing caller profiles, an association score based on the comparison. The method includes presenting data about the caller from the existing caller profile on a call-taking terminal of the service answering point assigned to the service call when the association score for one of the plurality of existing caller profiles exceeds an association threshold.