Domain name generation and searching using unigram queries

1. A system, comprising:

a server computer including at least one processor executing instructions that cause the server computer to:

identify, within a plurality of consecutive queries for domain names in at least one search log, a longest common string between each of the plurality of consecutive queries;

identify an interchangeable character string for the longest common string by searching the plurality of consecutive queries for a co-occurrence of at least one additional character string with the longest common string based on at least one frequency associated with at least one word pair within the plurality of consecutive queries;

identify at least one modification term of the at least one additional character string based on analysis of each additional character sting to determine a type associated with the interchangeable character string and the longest common string,

wherein the at least one modification term determines a relationship between the type associated with the interchangeable character string and the longest common string;

generate, for at least one query in the plurality of consecutive queries, at least one unique data record within a unigram query suggestion database, wherein the unique data record comprises:

the longest common string,

the interchangeable character string in association with the longest common string,

the modification term, and

a frequency of use based on a number of queries in the plurality of consecutive queries that comprise the interchangeable character string, and the longest common string;

receive, from a client computer, a new search string;

identify, within the new search string, a plurality of tokens based on the at least one unique data record by identifying a plurality of bigrams within a keyword dictionary database and applying a language model to determine at least acronym of the plurality of tokens,

wherein the plurality of tokens comprises a unigram matching the longest common string,

wherein at least one token of the plurality of tokens comprises at least one ngram capable of rewriting the unigram,

wherein the plurality of bigrams comprises at least two word pairs within the plurality of word pair types;

determine, based on the unigram query suggestion database, at least one domain name candidate comprising the at least one unique data record within the unigram and the interchangeable character string; and

transmit the domain name candidate to the client computer.

2. The system of claim 1 , wherein the instructions cause the server computer, responsive to the token within the new search string not being a unigram, to generate a plurality of alternative available domain name candidates according to an n-gram analysis.

3. The system of claim 2 , wherein the instructions cause the server computer to combine the at least one domain name candidate with the plurality of alternative

available domain name candidates.

4. The system of claim 3 , wherein the instructions cause the server computer to rank the at least one domain name candidate and the plurality of alternative

available domain name candidates.

5. The system of claim 1 , wherein the instructions cause the server computer to:

crawl at least one electronic document comprising a plurality of word pairs comprising interchangeable terms;

identify a rewrite distribution for a plurality of unigram queries within the electronic document;

generate an alternative available domain name candidate using at least one domain name spin technique;

store in association in the unigram query suggestion database:

the unigram;

the alternative available domain name candidate;

at least one feature of the alternative available domain name candidate; and

a ranking of each of the alternative available domain name candidate.

6. The system of claim 5 , wherein the at least one electronic document comprises a plurality of content input during a domain name search session.

7. The system of claim 5 , wherein the at least one domain name spin technique comprises a domain name spin according to a high frequency of rewritten domain name searches

including the unigram.

8. The system of claim 5 , wherein the at least one domain name spin technique comprises a domain name spin according to a high frequency of the token.

9. The system of claim 5 , wherein the unigram, the alternative available domain name candidate, the at least one feature, and the at least one ranking are stored offline.

10. A method comprising:

identifying, by a server computer and within a plurality of consecutive queries for domain names in at least one search log, a longest common string between each of the plurality of consecutive queries;

identifying, by the server computer, an interchangeable character string for the longest common string by searching the plurality of consecutive queries for a co-occurrence of at least one additional character string with the longest common string based on at least one frequency associated with at least one word pair within the plurality of consecutive queries;

identifying, by the server computer, at least one modification term of the at least one additional character string based on analysis of each additional character sting to determine a type associated with the interchangeable character string and the longest common string,

wherein the at least one modification term determines a type of a relationship between the interchangeable character string and the longest common string;

generating, by the server computer and for at least one query in the plurality of consecutive queries, a unique data record within at least one unigram query suggestion database,

wherein the unique data record comprising:

the longest common string,

the modification term;

the interchangeable character string in association with the longest common string, and

a frequency of use based on a number of queries in the plurality of consecutive queries that comprise the interchangeable character string, and the longest common string;

receiving, by the server computer, a new search string from a client computer;

identifying, by the sever computer and within the new search string, a plurality of tokens based on the at least one unique data record by identifying a plurality of bigrams within a keyword dictionary database and applying a language model to determine at least acronym of the plurality of tokens,

wherein the plurality of tokens comprises a unigram matching the longest common string,

wherein at least one token of the plurality of tokens comprises at least one ngram capable of rewriting the unigram,

wherein the plurality of bigrams comprises at least two word pairs within the plurality of word pair types;

determining, by the server computer and using the unigram query suggestion database, at least one domain name candidate comprising the at least one unique data record within the unigram and the interchangeable character string; and

transmitting, by the server computer, the domain name candidate to the client computer.

11. The method of claim 10 , further comprising the step of responsive to the token within the new search string not being a unigram, generating, by the server computer, a plurality of alternative available domain name candidates according to an n-gram analysis.

12. The method of claim 11 , further comprising the step of combining, by the server computer, the at least one domain name candidate with the plurality of alternative

available domain name candidates.

13. The method of claim 12 , further comprising the step of ranking, by the server computer, the at least one domain name candidate and the plurality of alternative available domain name candidates.

14. The method of claim 10 , further comprising the steps of:

crawling, by the server computer, at least one electronic document comprising a plurality of word pairs comprising interchangeable terms;

identifying, by the server computer, a rewrite distribution for a plurality of unigram queries within the electronic document;

generating, by the server computer, an alternative available domain name candidate using at least one domain name spin technique;

storing, by the server computer, in association in the unigram query suggestion database:

the unigram;

the alternative available domain name candidate;

at least one feature of the alternative available domain name candidate; and

a ranking of each of the alternative available domain name candidate.

15. The method of claim 14 , wherein the at least one electronic document comprises a plurality of content input during a domain name search session.

16. The method of claim 14 , wherein the at least one domain name spin technique comprises a domain name spin according to a high frequency of rewritten domain name searches including the unigram.

17. The method of claim 14 , wherein the at least one domain name spin technique comprises a domain name spin according to a high frequency of the

token.

18. The method of claim 14 , wherein the unigram, the alternative available domain name candidate, the at least one feature, and the at least one ranking are stored offline.

19. The computer-implemented method of claim 10 , further comprising:

crawling, by the server computer, at least one electronic document comprising a plurality of word pairs comprising interchangeable terms;

identifying, by the server computer, a rewrite distribution for a plurality of unigram queries within the electronic document;

generating, by the server computer, an alternative available domain name candidate using at least one domain name spin technique;

storing, by the server computer, in association in the unigram query suggestion database:

the unigram;

the alternative available domain name candidate;

at least one feature of the alternative available domain name candidate; and

a ranking of each of the alternative available domain name candidate.