Domain Name Classification Using ML for Infringement Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring domain names to prevent unauthorized use and infringement are computationally expensive and struggle to distinguish between random and actual domain names, especially for short or non-unique domains, leading to false positive matches.
Innovation Solution
A system and method using a machine learning model trained on semantically-meaningful words and random strings to classify domain names as random, indeterminate, or not-random by decomposing domain names into segments, calculating similarity scores, and applying a classification algorithm to identify potential infringement risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional domain monitoring methods are used to detect potential infringements, then domain security protection is achieved, but computational resources are excessively consumed and false positive rates increase
Solution Approach 1:
The domain name analysis process is segmented into multiple stages: initial filtering using n-gram matching, followed by machine learning classification for borderline cases. This segmentation allows the system to quickly process obvious matches while applying more computationally intensive methods only when necessary, thereby improving overall computational efficiency while maintaining detection reliability.
Solution Approach 2:
A machine learning classifier is introduced as an intermediary component between traditional keyword matching and final infringement determination. This intermediary evaluates ambiguous cases that don't clearly match or mismatch seed domains, reducing false positives by providing probabilistic assessments that guide further investigation or dismissal.
2Reliability
If traditional domain monitoring methods are used to detect potential infringements, then domain security protection is achieved, but false positive matches with random domain names increase
Solution Approach 1:
The system implements feedback mechanisms where classification results are continuously evaluated and used to refine the machine learning model. False positives from random domain names are fed back into the training process, allowing the model to learn patterns that distinguish legitimate domain variations from random strings, thereby improving classification precision over time while maintaining security protection.
Solution Approach 2:
Traditional mechanical keyword-matching systems are replaced with machine learning-based classification. Instead of relying on rigid string comparison algorithms that cannot distinguish between meaningful and random character sequences, the system uses trained models that have learned to recognize patterns of legitimate domain variations versus random noise, significantly improving measurement precision.
3Measurement precision
If machine learning models are trained on semantically-meaningful words and random strings, then classification accuracy improves, but training data requirements and model complexity increase
Solution Approach 1:
The system changes parameters of the machine learning model dynamically based on the specific analysis context. Different models or model configurations are applied depending on the domain characteristics, seed domain properties, and analysis stage. This allows the system to achieve high classification accuracy when needed while avoiding unnecessary model complexity for straightforward cases, effectively managing the trade-off between precision and complexity.
Data Source
AI summary
Aspects of the disclosure relate to detecting random and/or algorithmically-generated character sequences in domain names. A computing platform may train a machine learning model based on a set of semantically-meaningful words. Subsequently, the computing platform may receive a seed string and a set of domains to be analyzed in connection with the seed string. Based on the machine learning model, the computing platform may apply a classification algorithm to the seed string and the set of domains, where applying the classification algorithm to the seed string and the set of domains produces a classification result. Thereafter, the computing platform may store the classification result.