Voice Call Classification via Attribute Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current systems fail to effectively classify calls from malicious entities, including hacking calls with swapped numbers, leading to incorrect blocking of legitimate communications and inability to distinguish between malicious and legitimate calls.

Innovation Solution

A method for voice call classification that intercepts and analyzes voice call data, transforming it into a predefined format to identify attributes such as call category, emotional component, and presence of robotized speech, using machine learning algorithms like naive Bayesian classifiers and SVM to classify calls as regular or unwanted, and notify the user accordingly.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional systems block calls based on swapped numbers, then hacking attempts are partially blocked, but legitimate communications are incorrectly blocked and false positives increase

Engineering Contradiction:
Improvecall blocking accuracyVSAvoidfalse positives
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent introduces an intermediary classification system between the call blocking mechanism and the swapped number database. Instead of directly blocking calls with swapped numbers, the system first classifies them to determine if they are malicious or legitimate. This intermediary step prevents false positives by ensuring that only confirmed malicious calls are blocked, while legitimate calls with swapped numbers are allowed through.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system performs preliminary classification of calls before blocking them. By analyzing call attributes and comparing them against known patterns of malicious calls, the system determines the nature of the call in advance. This preliminary action ensures that legitimate communications are not incorrectly blocked, as they are identified and allowed through before the blocking mechanism is applied.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If database-based spam detection is used, then known spam numbers are blocked, but new hacking schemes with swapped numbers cannot be detected

Engineering Contradiction:
Improvespam detection accuracyVSAvoiddetection of new hacking schemes
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent implements a dynamic classification system that adapts to new hacking schemes. Instead of relying solely on a static database of known spam numbers, the system continuously analyzes call attributes and updates its classification models. This dynamic approach enables the system to detect new hacking schemes with swapped numbers by identifying patterns and anomalies in real-time, rather than requiring pre-programmed knowledge of every possible spam number.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the parameters used for detection from simple number matching to multi-attribute analysis. By examining multiple call attributes such as timing patterns, content analysis, and behavioral characteristics, the system can detect new hacking schemes even when the caller ID is swapped. This parameter change enables the system to adapt to evolving threats without requiring constant updates to the spam number database.

Inventive Principle:
Principle #35Parameter changes

3Object-affected harmful factors

If number swapping technology is used by hackers, then their identity is concealed, but legitimate subscribers are discredited when their numbers are blocked

Engineering Contradiction:
Improvehacker identity concealmentVSAvoidlegitimate subscriber reputation
Core Design Contradiction:
Object-affected harmful factorsVSReliability

Solution Approach 1:

The patent implements a feedback mechanism that provides users with information about why calls are blocked. When a call is blocked, the system communicates the classification reasoning to the user, allowing them to verify whether the block was correct. This feedback loop helps legitimate subscribers understand that their numbers are not being incorrectly blocked, as they can see the classification evidence and provide corrections if needed, thereby protecting their reputation.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The classification system acts as an intermediary between hackers using number swapping and the blocking mechanism. By introducing this intermediate classification layer, the system can distinguish between malicious and legitimate calls even when the caller ID is swapped. This protects legitimate subscribers from being discredited, as their calls are properly classified and allowed through, while hacker calls are blocked based on their actual content and patterns rather than their masked identities.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3873075A1System and method for call classification
Publication Date: 2021.09.01 AO KASPERSKY LAB
  • EP3873075A1 patent drawingFigure 1
  • EP3873075A1 patent drawingFigure 2
  • EP3873075A1 patent drawingFigure 3

AI summary

A method for voice call analysis and classification includes interceting a voice call session between an initiating device and a recipient device. Voice call data exchanged between the initiating device and the recipient device during the voice call session is transformed into a predefined data format. The transformed voice call data is analyzed to determine one or more attributes of the intercepted voice call. One or more features associated with the intercepted voice call session are identified based on the determined one or more attributes. The intercepted voice call is classified using the identified one or more features.