US 11,539,661 B2Granted

Using a learning algorithm to suggest domain names

AssigneeGo Daddy Operating Company, LLC

InventorsWei-Cheng Lai, Yu Tian, Wenbo Wang, Chungwei Yen

Life of this patent

Estimating the term…

Loading patent…

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.

Claims18

Dependent claims

A method for a learning algorithm to suggest a domain name to a user, comprising the steps of:

receiving, by a domain name registrar, in chronological order from oldest to newest, a first domain name search and a second domain name search from the user;
reading, by the domain name registrar, a previously registered domain name from a registrar customer account of the user;
tokenizing, by the domain name registrar, the second domain name search into a first plurality of character strings;
applying, by the domain name registrar, the previously registered domain name, the first domain name search and the first plurality of character strings to a corresponding plurality of cells in an encoder of a trained machine learning algorithm;
wherein the machine learning algorithm has been trained to obtain the trained machine learning algorithm, by:
utilizing training data comprising
an input comprising
a training domain name search, and
an output comprising
a training expected domain name suggestion;
applying the input to the encoder of the machine learning algorithm;
outputting a training actual domain name suggestion by a decoder of the machine learning algorithm;
comparing the training expected domain name suggestion to the training actual suggested domain name;
and
iteratively modifying the machine learning algorithm until a second training actual domain name suggestion is more likely to match the training expected domain name suggestion;
creating, by the encoder, a vector for the previously registered domain name, the first domain name search and the plurality of character strings;
applying, by the domain name registrar, the vector to the decoder of the trained machine learning algorithm;
creating, by the decoder, a second plurality of character strings;
combining, by the domain name registrar, the second plurality of character strings to create a suggested domain name;
transmitting, by the domain name registrar, the suggested domain name to the user;
receiving, by the domain name registrar, a selection of the suggested domain name for registration from the user;
and
registering, by the domain name registrar, the selected suggested domain name to the user.

20 more elements

A method for a learning algorithm to suggest a domain name to a user, comprising the steps of:

receiving, by a domain name registrar, in chronological order from oldest to newest, a first domain name search and a second domain name search from the user;
tokenizing, by the domain name registrar, the second domain name search into a first plurality of character strings;
applying, by the domain name registrar, the first domain name search and the first plurality of character strings to a corresponding plurality of cells in an encoder of a trained machine learning algorithm;
wherein the machine learning algorithm has been trained to obtain the trained machine learning algorithm, by:
utilizing training data comprising
an input comprising
a training domain name search, and
an output comprising
a training expected domain name suggestion;
applying the input to the encoder of the machine learning algorithm;
outputting a training actual domain name suggestion by a decoder of the machine learning algorithm;
comparing the training expected domain name suggestion to the training actual suggested domain name;
and
iteratively modifying the machine learning algorithm until a second training actual domain name suggestion is more likely to match the training expected domain name suggestion;
creating, by the encoder, a vector for the first domain name search and the first plurality of character strings;
applying, by the domain name registrar, the vector to the decoder of the trained machine learning algorithm;
creating, by the decoder, a second plurality of character strings;
combining, by the domain name registrar, the second plurality of character strings to create a suggested domain name;
transmitting, by the domain name registrar, the suggested domain name to the user;
receiving, by the domain name registrar, a selection of the suggested domain name for registration from the user;
and
registering, by the domain name registrar, the selected suggested domain name to the user.

19 more elements

A method for a learning algorithm to suggest a domain name to a user, comprising the steps of:

receiving, by a domain name registrar, a domain name search from the user;
reading, by the domain name registrar, a previously registered domain name from a registrar customer account of the user;
tokenizing, by the domain name registrar, the domain name search into a first plurality of character strings;
applying, by the domain name registrar, the previously registered domain name and the first plurality of character strings to a corresponding plurality of cells in an encoder of a trained machine learning algorithm;
wherein the machine learning algorithm has been trained to obtain the trained machine learning algorithm, by:
utilizing training data comprising
an input comprising
a training domain name search, and
an output comprising
a training expected domain name suggestion;
applying the input to the encoder of the machine learning algorithm;
outputting a training actual domain name suggestion by a decoder of the machine learning algorithm;
and
comparing the training expected domain name suggestion to the training actual suggested domain name;
and
iteratively modifying the machine learning algorithm until a second training actual domain name suggestion is more likely to match the training expected domain name suggestion;
creating, by the encoder, a vector for past registered domain names and the plurality of character strings;
applying, by the domain name registrar, the vector to the decoder of the trained machine learning algorithm;
creating, by the decoder, a second plurality of character strings;
combining, by the domain name registrar, the second plurality of character strings to create a suggested domain name;
transmitting, by the domain name registrar, the suggested domain name to the user;
receiving, by the domain name registrar, a selection of the suggested domain name for registration from the user;
and
registering, by the domain name registrar, the selected suggested domain name to the user.

21 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/259,808Filed 28 Jan 2019
Published: US 2020/0244622 A130 Jul 2020
Granted: 27 Dec 2022
Term ends: Not estimatedThe source records don’t support an estimate.

Classification20 codes

G06F 16/958
Retrieval from the web: organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
G06F 16/3347
Query execution: using vector based model
G06F 18/214
Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation: generating training…
G06N 3/0442
Recurrent networks, e.g. Hopfield networks: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]