Short-Message Text Classification Using Character-Category Frequencies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for detecting unsolicited text-based content, particularly in short messages, struggle with accuracy due to manipulation techniques like random words, misspellings, and contextual reliance, and are computationally inefficient.
Innovation Solution
A method involving character category sequencing and frequency distribution analysis is used to classify text-based content, reducing input dimensionality and computational complexity while maintaining accuracy through a machine learning model trained on predefined character categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional character sequence analysis is used for text classification, then detection accuracy can be maintained, but computational complexity and memory utilization increase significantly
Solution Approach 1:
The patent extracts only the essential categorical properties of characters (e.g., uppercase, lowercase, digit, punctuation) rather than analyzing complete character sequences. This extraction approach maintains classification accuracy by preserving meaningful patterns while eliminating redundant information, thereby reducing computational complexity and memory requirements.
Solution Approach 2:
The patent segments the text analysis process into character-level categorical identification rather than processing complete character sequences. By dividing the analysis into discrete categorical units and analyzing their frequency distributions, the system achieves efficient computation while maintaining detection accuracy.
2Measurement precision
If traditional character sequence analysis is used for text classification, then detection accuracy can be maintained, but memory utilization increases significantly
Solution Approach 1:
The patent extracts only the essential categorical properties of characters (e.g., uppercase, lowercase, digit, punctuation) rather than analyzing complete character sequences. This extraction approach maintains classification accuracy by preserving meaningful patterns while eliminating redundant information, thereby reducing computational complexity and memory requirements.
Solution Approach 2:
Instead of analyzing character sequences directly (traditional approach), the patent inverts the approach by analyzing the frequency distribution of character categories. This inversion transforms the problem from sequence matching to statistical analysis, significantly reducing memory requirements while maintaining classification effectiveness.
3Productivity
If character category sequences are used instead of character sequences, then computational efficiency improves, but detection accuracy may be affected
Solution Approach 1:
The patent changes the parameters of analysis from complete character sequences to categorical properties of characters. By transforming the input data into categorical frequencies (e.g., proportion of uppercase letters, digits, punctuation), the system achieves computational efficiency while preserving the essential linguistic patterns needed for accurate classification.
Solution Approach 2:
The patent applies different categorical properties to different aspects of character analysis (e.g., case sensitivity, character type). This local quality approach allows the system to capture meaningful linguistic features at the character level without requiring complete sequence analysis, thereby maintaining accuracy while improving efficiency.
Data Source
AI summary
Approaches to classifying text-based content are described herein. For example, a classification system performs operations that include receiving text-based content comprising a plurality of characters, generating a plurality of character category sequences using the plurality of characters and based on a plurality of predefined character categories, calculating a frequency distribution of the plurality of character category sequences, and classifying the text-based content based on the calculated frequency distribution. The classifying uses a machine learning model that has been trained using a plurality of examples of text-based content. Responsive to the classification, the system can take appropriate actions. For example, responsive to classifying the text-based content as unsolicited, the system can restrict distribution of the text-based content or generate an alert for the text-based content.