Spam Detection via Frequency Domain Transmission Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing spam message detection methods are ineffective in identifying spammers who evade detection by transmitting messages through multiple sources in turns, leading to high false negative rates and privacy concerns, and are complicated by the need for large data record systems and shared data across service operators.
Innovation Solution
A method and apparatus that collect and analyze the time domain transmission characteristics of message sources, computing frequency domain characteristics using techniques like Fourier transformation or Autoregressive modeling to identify spammers based on predefined conditions, thereby detecting spammers that use multiple message sources to transmit messages in turns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If message content based detection with preset keywords is used, then spam messages containing keywords can be detected, but detection speed is affected and false negative rate increases when keyword set is not optimally sized
Solution Approach 1:
The patent transforms the detection approach from time domain (keyword matching) to frequency domain (transmission characteristic analysis). By changing the parameter space from message content to transmission timing patterns, the system achieves both high accuracy and speed without the trade-off present in keyword-based methods
Solution Approach 2:
The patent replaces the mechanical keyword matching process with a signal processing approach using Fourier transformation. This substitution enables parallel processing of transmission characteristics, significantly improving detection speed while maintaining or enhancing detection accuracy
2Measurement precision
If transmission speed based detection is used to identify spammers sending bulk messages, then spammers transmitting large volumes can be detected, but spammers can evade detection by using multiple message sources in turns
Solution Approach 1:
The patent merges the transmission characteristics of multiple message sources by transforming them into the frequency domain. This combination reveals coordinated patterns that individual time-domain analyses would miss, enabling detection of spammer networks that rotate through multiple sources
Solution Approach 2:
The patent adds a frequency domain dimension to the analysis of message transmission characteristics. By examining patterns in the frequency domain rather than just time domain, the system can identify coordinated spamming behavior across multiple sources that appears random in the time domain
3Measurement precision
If user feedback based approach or social network archive is used, then spammer identification can be achieved, but large data record systems are required and complexity increases for sharing across service operators
Solution Approach 1:
The patent extracts only the essential transmission timing characteristics from message data, discarding the need for storing and processing large volumes of message content, user feedback, or social network archives. This extraction approach maintains detection accuracy while dramatically reducing system complexity and data storage requirements
Data Source
AI summary
A method, apparatus and computer program product for spam message detection. The method includes collecting time domain transmission characteristic of a message source; computing frequency domain transmission characteristic of the message source with the time domain transmission characteristic of the message source; and identifying the message source to be a spammer in response to the frequency domain transmission characteristic of the message source satisfying predefined criteria; wherein the steps of the method are carried out using a computer device. An apparatus and computer program product for carrying out the above method is also provided.