Characterizing data using descriptive tokens

Patent metadata

Field	Value
Publication number	US-10334072-B2
Application number	US-201514985230-A
Country	US
Kind code	B2
Filing date	Dec 30, 2015
Priority date	Dec 30, 2015
Publication date	Jun 25, 2019
Grant date	Jun 25, 2019

Abstract

In one embodiment, a method includes a computing device receiving postings from users of an online social networking system. A postings may include location data along with one or more tags that may describe the content of the posting. The computing device may identify regions and subregions from which the postings originated, and may determine a distribution of the tags according to two data dimensions: the ubiquity of the tags across the regions, and the ubiquity of the tags across the subregions. Based on the distribution, the computing device may create a neighborhood characterization to accurately describe one or more subregions. The computing device may also determine applications for the neighborhood characterization.

First claim

What is claimed is:

1. A method comprising:
by a computing device, receiving postings submitted by one or more users of a social-networking system, wherein a posting comprises location data and one or more tags;
by the computing device, for each of the postings, based on the location data, identifying one of a plurality of regions and one of a plurality of subregions within the regions;
by the computing device, determining a distribution of the tags according to two data dimensions, wherein the two data dimensions include a first one of the data dimensions comprising a degree of ubiquity of each tag across the regions and a second one of the data dimensions comprising a degree of ubiquity of the same tag across the subregions within the different regions;
by the computing device, based on the distribution, identifying one or more of the tags that are common to multiple ones of the subregions across the different regions with respect to information gains of the respective tags, wherein each information gain is a measure of difference between two probability distributions, wherein the two probability distributions includes a probability distribution that a posting occurs in a particular region with a tag and a probability distribution that the same posting originates from the particular region;
by the computing device, generating a neighborhood characterization based on the identified tags;
by the computing device, receiving one or more requests for characterizing particular items; and
by the computing device, applying the neighborhood characterization to the particular items.

2. The method of claim 1 , wherein a region comprises a state, province, county, or a city, and wherein a subregion comprises a neighborhood.

3. The method of claim 1 , further comprising:
generating a visual representation of the distribution of the tags.

4. The method of claim 1 , wherein generating a neighborhood characterization based on the identified tags comprises applying natural language processing to content associated with the postings to generate the neighborhood characterization.

5. The method of claim 1 , wherein applying the neighborhood characterization comprises sending sponsored content to one or more users of the social networking system that belong to a particular subregion, based on the success of the sponsored content in one or more different subregions with a similar degree of ubiquity of the tags across the subregions in the regions.

6. The method of claim 1 , wherein identifying one or more of the tags that are common to multiple ones of the subregions across the different regions comprises identifying tags that occur above a pre-determined rate among a plurality of users located within a single subregion.

7. The method of claim 1 , further comprising:
generating a visual representation of the neighborhood characterization.

8. The method of claim 1 , wherein the neighborhood characterization comprises a composite characterization of two or more different topics.

9. The method of claim 1 , further comprising generating one or more additional neighborhood characterizations based on the identified tags from one or more additional neighborhoods; and
quantifying the similarity between tags of different neighborhoods.

10. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
receive postings submitted by one or more users of a social-networking system, wherein a posting comprises location data and one or more tags;
for each of the postings, based on the location data, identify one of a plurality of regions and one of a plurality of subregions within the regions;
determine a distribution of the tags according to two data dimensions, wherein the two data dimensions include a first one of the data dimensions comprising a degree of ubiquity of each tag across the regions and a second one of the data dimensions comprising a degree of ubiquity of the same tag across the subregions within the different regions;
based on the distribution, identify one or more of the tags that are common to multiple ones of the subregions across the different regions with respect to information gains of the respective tags, wherein each information gain is a measure of difference between two probability distributions, wherein the two probability distributions includes a probability distribution that a posting occurs in a particular region with a tag and a probability distribution that a posting originates from the particular region;
generate a neighborhood characterization based on the identified tags;
receive one or more requests for characterizing particular items; and
apply the neighborhood characterization to the particular items.

11. The media of claim 10 , wherein a region comprises a state, province, county, or a city, and wherein a subregion comprises a neighborhood.

12. The media of claim 10 , wherein the software is further operable when executed to generate a visual representation of the distribution of the tags.

13. The media of claim 10 , wherein generating a neighborhood characterization based on the identified tags comprises applying natural language processing to content associated with the postings to generate the neighborhood characterization.

14. The media of claim 10 , wherein applying the neighborhood characterization comprises sending sponsored content to one or more users of the social networking system that belong to a particular subregion, based on the success of the sponsored content in one or more different subregions with a similar degree of ubiquity of the tags across the subregions in the regions.

15. The media of claim 10 , wherein identifying one or more of the tags that are common to multiple ones of the subregions across the different regions comprises identifying tags that occur above a pre-determined rate among a plurality of users located within a single subregion.

16. The media of claim 10 , wherein the software is further operable when executed to generate a visual representation of the neighborhood characterization.

17. The media of claim 10 , wherein the neighborhood characterization comprises a composite characterization of two or more different topics.

18. The media of claim 10 , wherein the software is further operable when executed to generate one or more additional neighborhood characterizations based on the identified tags from one or more additional neighborhoods; and
quantify the similarity between tags of different neighborhoods.

19. A system comprising: one or more processors; and a memory coupled to the processors comprising instructions executable by the processors, the processors being operable when executing the instructions to:
receive postings submitted by one or more users of a social-networking system, wherein a posting comprises location data and one or more tags;
for each of the postings, based on the location data, identify one of a plurality of regions and one of a plurality of subregions within the regions;
determine a distribution of the tags according to two data dimensions, wherein the two data dimensions include a first one of the data dimensions comprising a degree of ubiquity of each tag across the regions and a second one of the data dimensions comprising a degree of ubiquity of the same tag across the subregions within the different regions;
based on the distribution, identify one or more of the tags that are common to multiple ones of the subregions across the different regions with respect to information gains of the respec

Assignees

• Facebook Inc

Inventors

• Tannen Jonathan
• State Bogdan

Classifications

• H04L67/327

  Electricity · mapped topic

• G06F16/9537 Primary

  Spatial or temporal dependent retrieval, e.g. spatiotemporal queries · CPC title

• H04L67/306 Primary

  User profiles · CPC title

• H04L67/02

  based on web technology, e.g. hypertext transfer protocol [HTTP] · CPC title

• G06F16/35

  Clustering; Classification · CPC title