System and method for website categorization

• US 9,311,423 B1
• Filed: 02/13/2014
• Issued: 04/12/2016
• Est. Priority Date: 02/19/2010
• Status: Active Grant

• Current Assignee: Go Daddy Operating Company LLC (GoDaddy, Inc.)
• Original Assignee: Go Daddy Operating Company LLC (GoDaddy, Inc.)
• Inventors: Brown, Robert, Kamdar, Tapan, Lai, Wei-Cheng, Kirkish, Ryan, McLellan, Jeff
• Primary Examiner(s): Phillips, III, Albert
• Application Number:
  US14/180,249
• Time in Patent Office: 789 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0278 Product appraisal
• CPC Class Codes: H04L 61/5076 Update or notification mechanisms, e.g. DynDNS

Abstract

Systems and methods for the categorization of websites are presented. A website is categorized using one or a combination of its domain name and its web page content. The domain name is tokenized, and the tokens compared to categories in a category structure to determine probabilities that the token belongs to each category. Combinations of tokens are similarly compared to the categories. A category may be determined with reference to a vector space in which a training set of websites having known categories is converted according to a methodology into reference vectors containing keyword frequencies. A target website is converted to a target vector using the same methodology, and a distance score of the target vector to each reference vector is calculated. The website represented by the target vector is assigned the category of the reference vector having the lowest distance score.

Claims

1. A method, comprising:

receiving, by at least one server communicatively coupled to a network, one or more tokens together forming all or part of a string comprising a domain name;

comparing, by the at least one server, each of the one or more tokens to each of a plurality of categories in a category structure to determine, for each pairing of one of the tokens with one of the categories, a token probability that the token belongs to the category;

for one or more of the token probabilities, increasing or reducing the token probability according to a frequency at which the category associated with the token probability is selected as a correct category or declined as an incorrect category for the token associated with the token probability, the frequency identified from a plurality of domain name searches previously processed by a first of the at least one server;

calculating, by the at least one server from the token probabilities, a final probability of the string belonging to each category; and

categorizing, by the at least one server, the token in the category having the highest final probability.

2. The method of claim 1, wherein comparing each of the one or more tokens to each of the categories comprises applying a string matching algorithm to match the one or more tokens to the categories.

3. The method of claim 1, wherein each of the categories in the category structure comprises a category title and a set of keywords, and wherein comparing each of the one or more tokens to each of the categories comprises applying a string matching algorithm to match the tokens to the category title or to one or more keywords of the set of keywords.

4. The method of claim 3, further comprising collecting a plurality of keywords for the set of keywords of each category.

5. The method of claim 4, wherein one or more of the plurality of keywords for one or more of the categories is obtained from one or more domain names previously categorized in the category.

6. The method of claim 1, wherein calculating the final probability for each category comprises averaging, of all of the tokens, the token probabilities for the category.

7. The method of claim 6, wherein calculating the final probability for each category further comprises determining, from a supplemental data model, a coefficient for each token;

and modifying the token probability for each token by the token'"'"'s coefficient before averaging the token probabilities.

8. The method of claim 1, further comprising comparing, by the at least one server, each combination of two or more of the tokens to each category in the category structure to determine a joint probability that the combination of tokens belongs to the category, the final probability further comprising a compilation of the joint probabilities.

9. The method of claim 8, wherein calculating the final probability for each category comprises averaging together, of all of the tokens, the token probabilities and the joint probabilities for the category.

10. The method of claim 9, wherein calculating the final probability for each category further comprises determining, from a supplemental data model, a coefficient for each token and each combination of tokens and modifying the token probability for each token and the joint probability for each combination of tokens by the token'"'"'s coefficient before averaging the token probabilities.

11. A system, comprising:

at least one server computer in communication with a network, the at least one server computer including a processor configured to;

receive one or more tokens together forming all or part of a string comprising a domain name;

compare each of the one or more tokens to each of a plurality of categories in a category structure to determine, for each pairing of one of the tokens with one of the categories, a token probability that the token belongs to the category;

for one or more of the token probabilities, increase or reduce the token probability according to a first frequency at which the category associated with the token probability is selected as a correct category or declined as an incorrect category for the token associated with the token probability, the first frequency identified from a plurality of domain name searches previously processed by a first of the at least one server;

calculate, from the token probabilities for each of the plurality of categories, a final probability of the domain name belonging to the category; and

categorize the domain name in the category having the highest final probability.

12. The system of claim 11, wherein the processor is further configured to compare each combination of two or more of the tokens to each category in the category structure to determine, for each of the combinations as one of the token probabilities, a joint probability that the combination of tokens belongs to the category.

13. The system of claim 11, wherein the domain name is not related to an existing website.

14. The system of claim 11, wherein the processor is further configured to present a domain search interface to a user via a user device in electronic communication with the at least one server and to receive the domain name via the domain search interface.

15. The system of claim 14, wherein the processor is further configured to:

determine if the domain name is available for registration; and

responsive to a determination that the domain name is not available for registration, recommend to the user, via the domain search interface, one or more alternative domain names that are relevant to the category in which the domain name is categorized.

16. The system of claim 11, wherein, for each of the one or more tokens, the processor determines the token probability of the token for each of the plurality of categories by:

determining a second frequency of the token'"'"'s appearance in a set of keywords and keyword phrases relevant to the category;

calculating the token probability based on the second frequency;

determining whether the token appears in a title of the category; and

responsive to a determination that the token appears in the title of the category, increasing the token probability.

17. The system of claim 11, wherein the category structure comprises one or more hierarchical trees in which the categories are arranged into a plurality of tiers.

18. The system of claim 17, in which the category structure comprises North American Industry Classification System codes.