Aggregating Tags in Images

摘要

In one embodiment, a method includes receiving, from a first user associated with an online social network, a request to view an image associated with the online social network, accessing the image, wherein the image includes multiple tags corresponding to multiple objects, generating an aggregated tag for two or more of the tags based on the proximity of the tags to each other, determining a social-graph affinity for each object in the aggregated tag, and rendering the image with a tag-label for the aggregated tag, where the description in the tag-label is based on the social-graph affinity for the objects associated with the aggregated tag.

主权项

1. A method comprising, by a computing device:
receiving, from a first user associated with an online social network, a request to view an image associated with the online social network; accessing the image, wherein the image comprises a plurality of tags corresponding to a plurality of objects associated with the online social network, respectively, and wherein each tag is associated with a particular location within the image; generating an aggregated tag for two or more of the plurality of tags based on a distance of the locations of the tags to each other, wherein tags associated with locations within a threshold distance of each other are aggregated with the aggregated tag; determining a social-graph affinity for each object associated with the aggregated tag; and rendering the image with a tag-label for the aggregated tag, wherein the tag-label comprises a description of the objects associated with the aggregated tag, and wherein the description is based on the social-graph affinity for the objects associated with the aggregated tag.