Categorizing hash tags

• US 9,384,259 B2
• Filed: 02/03/2014
• Issued: 07/05/2016
• Est. Priority Date: 02/03/2014
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Wexler, Mike, Webb, Brian
• Primary Examiner(s): Vy, Hung T
• Application Number:
  US14/170,952
• Publication Number:
  US 20150220615A1
• Time in Patent Office: 883 Days
• Field of Search: 707/741, 707/738, 707/740, 707/747, 715/273, 715/771
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/2255 Hash tables
• CPC Class Codes: G06F 16/23 Updating
• CPC Class Codes: G06F 16/285 Clustering or classification

Abstract

A content item categorizer system retrieves content items from Internet sources. If a retrieved content item includes sufficient information for traditional categorization methods, then the system assigns one or more categories to the content item using such traditional methods. The system creates a metadata model, based on information about traditionally-categorized content items, that maps at least hashtags from the content items to one or more content categories. When the system retrieves a sparse-info item that does not include sufficient information for traditional categorization, the system applies the metadata model to categorize the content item using at least hashtags in the sparse-info item. The metadata model may also include information indicating mappings between categories and coincidence of hashtags and additional content item attributes. Also, the metadata model may provide information for categorizing sparse-info items based on multiple hashtags in the sparse-info item metadata.

Claims

1. A method comprising:

using a first categorizer, categorizing a set of objects into a plurality of categories;

wherein the set of objects includes a subset of categorized objects that are associated with a particular hashtag;

after the first categorizer has categorized the set of objects into the plurality of categories, generating a metadata model based, at least in part, on;

the categories to which the first categorizer has assigned each object of the set of objects, and

hashtags associated with each object of the set of objects;

wherein, the metadata model comprises mappings of hashtags to categories of the plurality of categories;

wherein, the mappings in the metadata model include one or more mappings that map the particular hashtag to one or more categories;

wherein the one or more mappings are created based on the categories to which the first categorizer has assigned objects in the subset of categorized objects;

obtaining a particular object that is associated with the particular hashtag; and

using a second categorizer, different than the first categorizer, to categorize the particular object based, at least in part, on the one or more mappings, in the metadata model, that map the particular hashtag to the one or more categories;

wherein the method is performed by one or more computing devices.

2. The method of claim 1, wherein the mappings in the metadata model include one or more mappings that map the particular hashtag to the one or more categories further comprises:

obtaining a particular object associated with the particular hashtag;

associating at least a particular category, of the one or more categories, with the particular object based, at least in part, on information included in the particular object other than the particular hashtag;

in response to associating the particular category with the particular object, modifying information for a particular mapping between the particular hashtag and the particular category.

3. The method of claim 1, wherein using a second categorizer, different than the first categorizer, to categorize the particular object comprises:

creating a categorization association between the particular object and a particular category of the one or more categories; and

establishing a particular level of confidence for the categorization association.

4. A method of claim 1 comprising:

identifying a plurality of categorized objects that are associated with a particular hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing one or more mappings between the particular hashtag and one or more categories;

wherein establishing one or more mappings between the particular hashtag and one or more categories further comprises;

identifying a first subset of the plurality of categorized objects that are associated with a first category of a plurality of related categories;

identifying a second subset of the plurality of categorized objects that are associated with a second category of the plurality of related categories;

wherein the first subset includes at least one categorized object that the second subset does not include;

determining that a third subset of the plurality of categorized objects, which includes the objects of both the first and second subsets, is associated with a category that represents the plurality of related categories; and

establishing a mapping between the particular hashtag and the category that represents the plurality of related categories;

obtaining a particular object associated with the particular hashtag; and

categorizing the particular object based, at least in part, on the one or more mappings between-the particular hashtag and the one or more categories;

wherein the method is performed by one or more computing devices.

5. The method of claim 1, further comprising:

using the first categorizer, categorizing a second set of objects into a second plurality of categories;

wherein the set of objects includes second subset of categorized objects that are associated with a second hashtag;

based, at least in part, on how the second subset of categorized objects have been categorized, generating a second mapping between the second hashtag and one or more certain categories;

wherein the particular object is further associated with the second hashtag; and

wherein using the second categorizer to categorizes the particular object comprises categorizing the particular object based, at least in part, on the second mapping.

6. The method of claim 1, further comprising:

using the first categorizer, categorizing a subset of the set of objects that are associated with both the particular hashtag and a second hashtag;

based, at least in part, on how the subset of the set of objects have been categorized, generating a second mapping between one or more certain categories and coincidence of the particular hashtag and the second hashtag;

wherein the particular object is further associated with the second hashtag; and

wherein using a second categorizer to categorize the particular object comprises categorizing the particular object based, at least in part, on the second mapping.

7. A method comprising:

identifying a plurality of categorized objects that are associated with a particular hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing one or more mappings between the particular hashtag and one or more categories;

obtaining a particular object associated with the particular hashtag; and

categorizing the particular object based, at least in part, on the one or more mappings between-the particular hashtag and the one or more categories;

identifying a particular value of a particular object attribute other than a hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing a second mapping between one or more certain categories and coincidence of the particular hashtag and the particular value of the particular object attribute;

obtaining a second object associated with the particular hashtag; and

categorizing the second object based, at least in part, on the second mapping

wherein the method is performed by one or more computing devices.

8. The method of claim 7, wherein:

the one or more mappings between the particular hashtag and one or more categories includes a first mapping that maps a particular category with the particular hashtag; and

the second mapping maps the particular category with coincidence of the particular hashtag and the particular value of the particular object attribute;

a level of confidence for the second mapping is higher than a level of confidence for the first mapping; and

categorizing the second object creates a categorization association between the second object and the particular category with a particular level of confidence based on the level of confidence for the second mapping.

9. The method of claim 7, further comprising:

determining that a level of confidence for at least a first mapping of the one or more mappings is below a particular threshold; and

in response to determining that the level of confidence for at least the first mapping is below the particular threshold, performing a correlation analysis for the plurality of categorized objects that are associated with the particular hashtag;

wherein performing the correlation analysis includes identifying a particular value of a particular object attribute other than a hashtag.

10. A non-transitory computer-readable medium that stores instructions which, when executed by one or more processors, cause performance of:

using a first categorizer, categorizing a set of objects into a plurality of categories;

wherein the set of objects includes a subset of categorized objects that are associated with a particular hashtag;

after the first categorizer has categorized the set of objects into the plurality of categories, generating a metadata model based, at least in part, on;

the categories to which the first categorizer has assigned each object of the set of objects, and

hashtags associated with each object of the set of objects;

wherein, the metadata model comprises mappings of hashtags to categories of the plurality of categories;

wherein, the mappings in the metadata model include one or more mappings that map the particular hashtag to one or more categories;

wherein the one or more mappings are created based on the categories to which the first categorizer has assigned objects in the subset of categorized objects;

obtaining a particular object that is associated with the particular hashtag; and

using a second categorizer, different than the first categorizer, to categorize the particular object based, at least in part, on the one or more mappings, in the metadata model, that map the particular hashtag to the one or more categories.

11. The non-transitory computer-readable medium of claim 10, wherein the mappings in the metadata model include one or more mappings that map the particular hashtag to the one or more categories further comprises:

obtaining a particular object associated with the particular hashtag;

associating at least a particular category, of the one or more categories, with the particular object based, at least in part, on information included in the particular object other than the particular hashtag;

in response to associating the particular category with the particular object, modifying information for a particular mapping between the particular hashtag and the particular category.

12. The non-transitory computer-readable medium of claim 10, wherein using a second categorizer, different than the first categorizer, to categorize the particular object comprises:

creating a categorization association between the particular object and a particular category of the one or more categories; and

establishing a particular level of confidence for the categorization association.

13. The non-transitory computer-readable medium of claim 10, further comprising instructions for:

using the first categorizer, categorizing a second set of objects into a second plurality of categories;

wherein the set of objects includes second subset of categorized objects that are associated with a second hashtag;

based, at least in part, on how the second subset of categorized objects have been categorized, generating a second mapping between the second hashtag and one or more certain categories;

wherein the particular object is further associated with the second hashtag; and

wherein using the second categorizer to categorizes the particular object comprises categorizing the particular object based, at least in part, on the second mapping.

14. The non-transitory computer-readable medium of claim 10, further comprising instructions for:

using the first categorizer, categorizing a subset of the set of objects that are associated with both the particular hashtag and a second hashtag;

based, at least in part, on how the subset of the set of objects have been categorized, generating a second mapping between one or more certain categories and coincidence of the particular hashtag and the second hashtag;

wherein the particular object is further associated with the second hashtag; and

wherein using a second categorizer to categorize the particular object comprises categorizing the particular object based, at least in part, on the second mapping.

15. A non-transitory computer-readable medium that stores instructions which, when executed by one or more processors, cause performance of:

identifying a plurality of categorized objects that are associated with a particular hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing one or more mappings between the particular hashtag and one or more categories;

obtaining a particular object associated with the particular hashtag; and

categorizing the particular object based, at least in part, on the one or more mappings between-the particular hashtag and the one or more categories;

identifying a particular value of a particular object attribute other than a hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing a second mapping between one or more certain categories and coincidence of the particular hashtag and the particular value of the particular object attribute;

obtaining a second object associated with the particular hashtag; and

categorizing the second object based, at least in part, on the second mapping.

16. The non-transitory computer-readable medium of claim 15, wherein:

the one or more mappings between the particular hashtag and one or more categories includes a first mapping that maps a particular category with the particular hashtag; and

the second mapping maps the particular category with coincidence of the particular hashtag and the particular value of the particular object attribute;

a level of confidence for the second mapping is higher than a level of confidence for the first mapping; and

categorizing the second object creates a categorization association between the second object and the particular category with a particular level of confidence based on the level of confidence for the second mapping.

17. The non-transitory computer-readable medium of claim 15, further comprising instructions for:

determining that a level of confidence for at least a first mapping of the one or more mappings is below a particular threshold; and

in response to determining that the level of confidence for at least the first mapping is below the particular threshold, performing a correlation analysis for the plurality of categorized objects that are associated with the particular hashtag;

wherein performing the correlation analysis includes identifying a particular value of a particular object attribute other than a hashtag.

18. A non-transitory computer-readable medium that stores instructions which, when executed by one or more processors, cause performance of:

identifying a plurality of categorized objects that are associated with a particular hashtag;

based, at least in part, on how the plurality of categorized objects have been categorized, establishing one or more mappings between the particular hashtag and one or more categories;

wherein establishing one or more mappings between the particular hashtag and one or more categories further comprises;

identifying a first subset of the plurality of categorized objects that are associated with a first category of a plurality of related categories;

identifying a second subset of the plurality of categorized objects that are associated with a second category of the plurality of related categories;

wherein the first subset includes at least one categorized object that the second subset does not include;

determining that a third subset of the plurality of categorized objects, which includes the objects of both the first and second subsets, is associated with a category that represents the plurality of related categories; and

establishing a mapping between the particular hashtag and the category that represents the plurality of related categories;

obtaining a particular object associated with the particular hashtag; and

categorizing the particular object based, at least in part, on the one or more mappings between-the particular hashtag and the one or more categories.