Creating training data for a learning algorithm to suggest domain names

Abstract

Methods are taught for creating training data for a learning algorithm, training the learning algorithm with the training data and using the trained learning algorithm to suggest domain names to users. A domain name registrar may store activities of a user on a registrar website. Preferably, domain name searches, selected suggested domain names and domain names registered to the user are stored as the training data in a training database. The training data may be stored so that earlier activities act as inputs to the learning algorithm while later activities are the expected outputs of the learning algorithm. Once trained, the learning algorithm may receive activities of other users and suggest domain names to the other users based on their activities.

The invention claimed is: 
     
         1 . A method for creating training data for a learning algorithm to suggest domain names, comprising the steps of:
 receiving, by a domain name registrar, from a first user in a single continuous account logon session, in chronological order from oldest to newest, a first domain name search, a second domain name search and a third domain name search;   generating, by the domain name registrar, a plurality of suggested domain names based on the first domain name search, the second domain name search or the third domain name search;   transmitting, by the domain name registrar, the plurality of suggested domain names to the first user;   receiving, by the domain name registrar, from the first user, a first selected domain name in the plurality of suggested domain names to add to an electronic shopping cart and a second selected domain name in the plurality of suggested domain names to add to the electronic shopping cart;   registering, by the domain name registrar, either the first selected domain name or the second selected domain name to the first user, thereby creating a registered domain name;   generating, by the domain name registrar, training data, comprising a first plurality of pairs, for the learning algorithm,
 wherein a first pair, in the first plurality of pairs, comprises the first domain name search as an input domain search and the second domain name search as an expected domain name suggestion, and 
 wherein a second pair, in the first plurality of pairs, comprises the second domain name search as an input domain search and the third domain name search as an expected domain name suggestion; and 
   storing, by the domain name registrar, the training data in a training database.   
     
     
         2 . The method of  claim 1 , further comprising the steps of:
 tokenizing the first domain name search, the second domain name search and the third domain name search into one or more strings of characters; and   confirming at least one string of characters, other than a top-level domain, in the first domain name search, the second domain name search and the third domain name search all match or are synonyms with each other.   
     
     
         3 . The method of  claim 1 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the second domain name search as an input domain search and the first domain name search as an expected domain name suggestion. 
     
     
         4 . The method of  claim 1 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the first selected domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         5 . The method of  claim 1 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the registered domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         6 . The method of  claim 1 , wherein the training data further comprises a second plurality of pairs generated from user activity of a second user in a single continuous account logon session of the second user,
 wherein a first pair, in the second plurality of pairs from the second user, comprises a domain name search from the second user as an input domain search and a registered domain name registered to the second user as an expected domain name suggestion.   
     
     
         7 . A method for creating training data for a learning algorithm to receive a domain name search and then suggest a domain name, comprising the steps of:
 receiving, by a domain name registrar, from a first user in a single continuous account logon session, in chronological order from oldest to newest, a first domain name search, a second domain name search and a third domain name search;   generating, by the domain name registrar, a plurality of suggested domain names based on the first domain name search, the second domain name search or the third domain name search;   transmitting, by the domain name registrar, the plurality of suggested domain names to the first user;   receiving, by the domain name registrar, from the first user, a first selected domain name in the plurality of suggested domain names to add to an electronic shopping cart and a second selected domain name in the plurality of suggested domain names to add to the electronic shopping cart;   registering, by the domain name registrar, either the first selected domain name or the second selected domain name to the first user, thereby creating a registered domain name;   generating, by the domain name registrar, training data, comprising a first plurality of pairs, for the learning algorithm,
 wherein a first pair, in the first plurality of pairs, comprises the first domain name search as an input domain search and the first selected domain name as an expected domain name suggestion, and 
 wherein a second pair, in the first plurality of pairs, comprises the first domain name search as an input domain search and the second selected domain name as an expected domain name suggestion; and 
   storing, by the domain name registrar, the training data in a training database.   
     
     
         8 . The method of  claim 7 , further comprising the steps of:
 tokenizing the first domain name search, the first selected domain name and the second selected domain name into one or more strings of characters; and   confirming at least one string of characters, other than a top-level domain, in the first domain name search, the first selected domain name and the second selected domain name all match or are synonyms with each other.   
     
     
         9 . The method of  claim 7 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the second domain name search as an input domain search and the first domain name search as an expected domain name suggestion. 
     
     
         10 . The method of  claim 7 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the first selected domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         11 . The method of  claim 7 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the registered domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         12 . The method of  claim 7 , wherein the training data further comprises a second plurality of pairs generated from user activity of a second user in a single continuous account logon session of the second user,
 wherein a first pair, in the second plurality of pairs from the second user, comprises a domain name search from the second user as an input domain search and a registered domain name registered to the second user as an expected domain name suggestion.   
     
     
         13 . A method for creating training data for a learning algorithm to receive a domain name search and then suggest a domain name, comprising the steps of:
 receiving, by a domain name registrar, from a first user in a single continuous account logon session, in chronological order from oldest to newest, a first domain name search, a second domain name search and a third domain name search;   generating, by the domain name registrar, a plurality of suggested domain names based on the first domain name search, the second domain name search or the third domain name search;   transmitting, by the domain name registrar, the plurality of suggested domain names to the first user;   receiving, by the domain name registrar, from the first user, a first selected domain name in the plurality of suggested domain names to add to an electronic shopping cart and a second selected domain name in the plurality of suggested domain names to add to the electronic shopping cart;   registering, by the domain name registrar, either the first selected domain name or the second selected domain name to the first user, thereby creating a registered domain name;   generating, by the domain name registrar, training data, comprising a first plurality of pairs, for the learning algorithm,
 wherein a first pair, in the first plurality of pairs, comprises the first domain name search as an input domain search and the registered domain name as an expected domain name suggestion; and 
   storing, by the domain name registrar, the training data in a training database.   
     
     
         14 . The method of  claim 13 , further comprising the steps of:
 tokenizing the first domain name search, the second domain name search and the third domain name search into one or more strings of characters; and   confirming at least one string of characters, other than a top-level domain, in the first domain name search and the registered domain name match or are synonyms with each other.   
     
     
         15 . The method of  claim 13 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the second domain name search as an input domain search and the first domain name search as an expected domain name suggestion. 
     
     
         16 . The method of  claim 13 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the first selected domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         17 . The method of  claim 13 , wherein the training data, comprising the first plurality of pairs, does not include a pair that comprises the registered domain name as an input domain name search and the first domain name search as an expected domain name suggestion. 
     
     
         18 . The method of  claim 13 , wherein the training data further comprises a second plurality of pairs generated from user activity of a second user in a single continuous account logon session of the second user,
 wherein a first pair, in the second plurality of pairs from the second user, comprises a domain name search from the second user as an input domain search and a registered domain name registered to the second user as an expected domain name suggestion.