Methods and system for associating locations with annotations

• US 10,068,178 B2
• Filed: 01/13/2010
• Issued: 09/04/2018
• Est. Priority Date: 01/13/2010
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Oath Inc. (Verizon Communications Inc.)
• Inventors: van Zwol, Roelof, Murdock, Vanessa, Serdyukov, Pavel
• Primary Examiner(s): Thomas, Ashish K
• Assistant Examiner(s): Ohba, Mellissa M
• Application Number:
  US12/686,883
• Publication Number:
  US 20110173150A1
• Time in Patent Office: 3,156 Days
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Methods, systems and computer program products for associating geographical locations with annotations corresponding to content. In one method, a language model is developed. The language model is developed from the location information and the one or more annotations associated with content uploaded by users. The language model is based on the probabilistic distribution of locations over one or more annotations. Further, when a user provides one or more annotations, the system and the method may use the language model to identify one or more locations associated with the one or more annotations provided by the user. The language model predicts one or more geographical locations based on the probabilistic distribution of locations over the annotations.

Claims

1. A method comprising:

accessing, using one or more computing systems, one or more geographical locations and one or more first annotations for one or more items of content, wherein the one or more first annotations comprise at least one designated tag;

determining a probability distribution of one or more geographical locations over the one or more first annotations;

smoothing the probability distribution based, at least in part, on annotation-specific smoothing with the one or more items of content based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is proportional to a spatial ambiguity of the annotation;

storing the smoothed probability distribution in a language model data store; and

associating the one or more first annotations with one or more cells of a map having a cell grid layout indicating corresponding geographical locations based, at least in part, on the smoothed probability distribution;

wherein at least partially in response to receiving one or more second annotations, determining one or more of the geographical locations corresponding to the one or more second annotations based, at least in part, on the smoothed probability distribution.

2. The method of claim 1, wherein the one or more geographical locations comprise latitude and longitude coordinates of the at least one geographical location.

3. The method of claim 1, wherein the probability distribution comprises a multinomial probability distribution.

4. The method of claim 1, wherein the smoothing of the probability distribution is based, at least in part, on the annotation-based smoothing with neighboring locations within a predefined distance of one another.

5. The method of claim 1, further comprising smoothing the probability distribution based, at least in part, on cell based smoothing with neighboring, locations within a predefined distance of one another.

6. The method of claim 1, further comprising boosting the one or more first annotations based, at least in part, on an external knowledge base of locations.

7. The method of claim 1, wherein the one or more first or second annotations include at least one of:

a tag, a title, a description, or a comment.

8. The method of claim 1, wherein the one or more items of content include at least one of:

a multimedia content, a visual content, an audio content, or a text content.

9. A method comprising:

receiving, at a computing system, an item of content from a user, the item of content comprising one or more first annotations, wherein the one or more first annotations comprise at least one designated tag;

identifying a probability distribution of the one or more first annotations over two or more geographic locations based, at least in part, on a language model;

smoothing the probability distribution based, at least in part, on annotation-specific smoothing with the item of content based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is proportional to a spatial ambiguity of the annotation;

receiving one or more second annotations; and

wherein in response to receiving the one or more second annotations, determining one or more of the geographical locations corresponding to the one or more second annotations based, at least in part, on the smoothed probability distribution.

10. The method of claim 9, wherein the probability distribution of locations over the one or more first annotations comprises a multinomial probability distribution.

11. The method of claim 9, further comprising ranking the two or more geographical locations.

12. The method of claim 9, wherein the one or more first or second annotations include at least one of:

a tag, a title, a description, a keyword of a query, or a comment.

13. The method of claim 9, wherein the one or more first or second annotations are associated with at least one of:

a multimedia content, a visual content, an audio content, or a text content.

14. The method of claim 9, wherein the smoothing the probability distribution is further based, at least in part, on cell based smoothing with neighbors, wherein locations within a specific distance of one another are considered as neighbors.

15. The method of claim 9, further comprising boosting the one or more first or second annotations based, at least in part, on an external knowledge base of locations.

16. A system comprising:

at least one processor;

a memory; and

computer program code stored in a non-transitory computer readable storage medium, wherein the computer program code is executable by the at least one processor to;

access, using one or more computing systems, one or more geographical locations and one or more first annotations for one or more items of content, wherein the one or more first annotations are to comprise at least one tag to be designated;

determine a probability distribution of one or more geographical locations over the one or more first annotations;

smooth the probability distribution to be based, at least in part, on annotation-specific smoothing with the one or more items of content to be based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is to be proportional to a spatial ambiguity of the annotation;

store the smoothed probability distribution in a language model data store; and

associate the one or more first annotations with one or more cells of a map to have a cell grid layout to correspond to geographical locations to be based, at least in part, on the smoothed probability distribution;

wherein at least partially in response to receipt of one or more second annotations, to determine one or more of the geographical locations to correspond to the one or more second annotations to be based, at least in part, on the smoothed probability distribution.

17. The system of claim 16, the one or more first or second annotations to include at least one of a tag, a title, a description, or a comment.

18. The system of claim 16, the one or more items of content to include at least one of a multimedia content, a visual content, an audio content, or a text content.

19. A system comprising:

at least one processor;

a memory; and

computer program code stored in a non-transitory computer readable storage medium, wherein the computer program code is executable by the at least one processor to;

process an item of content from a user, the item of content to comprise one or more first annotations, wherein the one or more first annotations are to comprise at least one tag to be designated;

identify a probability distribution of the one or more first annotations over two or more geographic locations to be based, at least in part, on a language model;

smoothing the probability distribution to be based, at least in part, on annotation-specific smoothing with the item of content to be based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is to be proportional to a spatial ambiguity of the annotation;

process one or more second annotations to be received; and

wherein in response to receipt of the one or more second annotations, determine one or more of the geographical locations to correspond to the one or more second annotations to be based, at least in part, on the smoothed probability distribution.

20. The system of claim 19, the one or more first or second annotations to include at least one of a tag, a title, a description, a keyword of a query, or a comment.

21. The system of claim 19, the one or more annotations are to be associated with the one or more items of content, the one or more items of content to include at least one of a multimedia content, a visual content, an audio content, or a text content.

22. A non-transitory computer readable storage medium storing thereon computer-readable instructions, the computer-readable instructions being executable by one or more processors to:

access, using one or more computing systems, one or more geographical locations and one or more first annotations for one or more items of content, wherein the one or more first annotations comprise at least one designated tag;

determine a probability distribution of one or more geographical locations over the one or more first annotations;

smooth the probability distribution to be based, at least in part, on annotation-specific smoothing with the one or more items of content to be based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is to be proportional to a spatial ambiguity of the annotation;

store the smoothed probability distribution in a language model data store; and

associate the one or more first annotations with one or more cells of a map to have a cell grid layout to indicate corresponding geographical locations to be based, at least in part, on the smoothed probability distribution;

wherein at least partially in response to receipt of one or more second annotations, determine one or more of the geographical locations to correspond to the one or more second annotations to be based, at least in part, on the smoothed probability distribution.

23. A non-transitory computer readable storage medium storing thereon computer-readable instructions, the computer-readable instructions being executable by one or more processors to:

process an item of content from a user, the item of content to comprise one or more first annotations, wherein the one or more first annotations are to comprise at least one tag to be designated;

identify a probability distribution of the one or more first annotations over two or more geographic locations, to be based, at least in part, on a language model;

smooth the probability distribution to be based, at least in part, on annotation-specific smoothing with the item of content to be based, at least in part on a smoothing coefficient, wherein the smoothing coefficient for a particular annotation is to be proportional to a spatial ambiguity of the annotation;

process one or more second annotations to be received; and

wherein in response to receipt of the one or more second annotations, to determine one or more of the geographical locations to correspond to the one or more second annotations to be based, at least in part, on the smoothed probability distribution.