Machine learning-based generation of similar domain names
A method of generating receiving a valid domain name comprises evaluating a received valid domain name in a neural network trained to generate similar domain names, and providing an output comprising at least one domain name similar to the received valid domain name generated by the neural network. In a further example, a recurrent neural network is trained using valid domain names and observed malicious similar domain names and/or linguistic rules. In another example, the output of the recurrent neural network further comprises a similarity score reflecting a degree of similarity between the valid domain name and the similar domain name, such that the similarity score can be used to generate a ranked list of domain names similar to the valid domain name.
1. A method of generating similar domain names using machine learning, comprising:
receiving a valid domain name;
evaluating the received valid domain name in a neural network trained to generate similar domain names, wherein the neural network comprises a recurrent neural network using a bidirectional encoder and a unidirectional encoder; and
providing an output comprising at least one domain name similar to the received valid domain name generated by the neural network.
2. The method of generating similar domain names using machine learning of claim 1 , wherein the neural network is trained using valid domain names and observed malicious similar domain names.
3. The method of generating similar domain names using machine learning of claim 1 , wherein the neural network is trained using linguistic rules.
4. The method of generating similar domain names using machine learning of claim 1 , wherein providing an output further comprises providing a ranked list of domain names similar to the received valid domain name.
5. The method of generating similar domain names using machine learning of claim 1 , further comprising registering at least one of the at least one output domain names similar to the received valid domain name.
6. The method of generating similar domain names using machine learning of claim 1 , further comprising indicating which of the at least one output domain names are registered.
7. The method of generating similar domain names using machine learning of claim 1 , wherein the output of the neural network further comprises a similarity score reflecting a degree of similarity between the valid domain name and the similar domain name.
8. The method of generating similar domain names using machine learning of claim 7 , further comprising using the similarity score of two or more generated similar domain names to generate a ranked list of domain names similar to the valid domain name.
9. A method of training a neural network to generate similar domain names, comprising:
receiving a valid domain name and a similar domain name;
evaluating the received valid domain name in a neural network and observing an output of the neural network, wherein the neural network comprises a bidirectional encoder and a unidirectional encoder; and
if the similar domain name and the observed output of the neural network differ, modifying at least one parameter of the neural network to generate an output more similar to similar domain name than the observed output of the neural network.
10. The method of training a neural network to generate similar domain names of claim 9 , further comprising using linguistic rules to train the neural network.
11. The method of training a neural network to generate similar domain names of claim 9 , wherein the output of the neural network further comprises a similarity score reflecting a degree of similarity between the valid domain name and the similar domain name.
12. The method of training a neural network to generate similar domain names of claim 9 , wherein the neural network comprises a plurality of Long Short Term Memory (LSTM) nodes.
13. The method of training a neural network to generate similar domain names of claim 9 , wherein the neural network comprises a plurality of Gated Recurrent Units (GRUs).
14. A method of generating similar domain names using machine learning, comprising:
receiving a valid domain name;
evaluating the received valid domain name in a machine learning system trained to generate similar domain names, wherein the machine learning system comprises a recurrent neural network using a bidirectional encoder and a unidirectional encoder; and
providing an output comprising at least one domain name similar to the received valid domain name generated by the machine learning system.
15. The method of generating similar domain names using machine learning of claim 14 , wherein the machine learning system is trained using valid domain names and observed malicious similar domain names.
16. The method of generating similar domain names using machine learning of claim 14 , wherein the machine learning is trained using linguistic rules.
17. The method of generating similar domain names using machine learning of claim 14 , wherein providing an output further comprises providing a ranked list of domain names similar to the received valid domain name.
18. The method of generating similar domain names using machine learning of claim 14 , wherein the output of the machine learning system further comprises a similarity score reflecting a degree of similarity between the valid domain name and the similar domain name.