Including content created by an online system user in a page associated with an entity and/or adding data to the content based on a measure of quality of an image included in the content

Abstract

An online system receives a content item including an image from an online system user. The online system accesses and applies a trained item detection model to predict a probability that a region of interest within the image corresponds to an item associated with an entity based on a set of pixel values associated with the region of interest. If the probability is at least a threshold probability, the online system accesses and applies a trained quality prediction model to predict a measure of quality of the image based on a set of attributes of the image. If the measure of quality is at least a threshold measure of quality, the online system includes the content item in a page associated with the entity maintained in the online system and/or adds a set of data associated with the item and/or the entity to the content item.

What is claimed is: 
     
         1 . A method comprising:
 receiving a content item comprising an image from a content-providing user of an online system;   identifying a first region of interest within the image that depicts a first object, wherein the first object corresponds to a first item associated with a first entity having a presence on the online system;   identifying a second region of interest within the image that depicts a second object, wherein the second object corresponds to a second item associated with a second entity having a presence on the online system;   predicting a first measure of quality of the image by applying a quality prediction model to the image, wherein the quality prediction model is a machine-learning model trained to predict a measure of quality of an image based on the image and a set of attributes of the image;   determining that the first measure of quality exceeds a threshold measure of quality;   responsive to determining that the first measure of quality does not exceed the threshold measure of quality, modifying the image to remove the second region of interest;   predicting a second measure of quality of the modified image by applying the quality prediction model to the modified image;   determining that the second measure of quality exceeds the threshold measure of quality; and   responsive to determining that the second measure of quality exceeds the threshold measure of quality, adding a set of data to the content item, wherein the set of data is associated with the first item and the first entity.   
     
     
         2 . The method of  claim 1 , wherein the set of data added to the content item comprises one or more selected from the group consisting of: a tag describing the item, a link to the page associated with the entity maintained in the online system, and a catalog of items associated with the entity maintained in the online system. 
     
     
         3 . The method of  claim 1 , further comprising:
 responsive to determining that the first measure of quality does not exceed the threshold, one or more selected from the group consisting of: cropping a portion of the image, zooming into a portion of the image, increasing a sharpness of the image, reducing an amount of noise characterizing the image, changing one or more colors within the image, changing a brightness of the image, and reducing an amount of distortion characterizing the image.   
     
     
         4 . The method of  claim 1 , wherein the quality prediction model is trained based on an additional set of images with different measures of quality. 
     
     
         5 . The method of  claim 4 , wherein the different measures of quality of the additional set of images are based at least in part on one or more selected from the group consisting of: a resolution of each of the additional set of images, a percentage of each of the additional set of images corresponding to an item associated with the entity, a number of items associated with one or more additional entities having a presence on the online system comprising each of the additional set of images, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises a content item comprising each of the additional set of images. 
     
     
         6 . The method of  claim 1 , wherein the predicted first measure of quality of the image is based at least in part on one or more selected from the group consisting of: a resolution of the image, a percentage of the image corresponding to the item associated with the entity, a number of items associated with the one or more additional entities having a presence on the online system comprising the image, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         7 . The method of  claim 6 , wherein the predicted first measure of quality is proportional to one or more of: the resolution of the image, the percentage of the image corresponding to the item associated with the entity, and the amount of user engagement with the link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         8 . The method of  claim 6 , wherein the predicted first measure of quality is inversely proportional to the number of items associated with the one or more additional entities having a presence on the online system comprising the image. 
     
     
         9 . The method of  claim 1 , wherein the predicted second measure of quality of the image is based at least in part on one or more selected from the group consisting of: a resolution of the image, a percentage of the image corresponding to the item associated with the entity, a number of items associated with the one or more additional entities having a presence on the online system comprising the image, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         10 . The method of  claim 1 , further comprising:
 responsive to determining that the second predicted measure of quality exceeds the threshold measure of quality:
 identifying one or more viewing users of the online system subscribing to content received from the content-providing user, wherein the one or more viewing users satisfy a set of targeting criteria received from the first entity; 
 determining a number of the one or more viewing users subscribing to the content received from the content-providing user; 
 determining a value of a performance metric associated with one or more content items received from the content-providing user, wherein the one or more content items are associated with one or more of:
 the one or more items associated with the first entity and a topic associated with the one or more items; and 
 
 computing an influencer score associated with the content-providing user based at least in part on one or more of: the number of the one or more viewing users subscribing to the content received from the content-providing user and the value of the performance metric. 
   
     
     
         11 . A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to:
 receive a content item comprising an image from a content-providing user of an online system;   identify a first region of interest within the image that depicts a first object, wherein the first object corresponds to a first item associated with a first entity having a presence on the online system;   identify a second region of interest within the image that depicts a second object, wherein the second object corresponds to a second item associated with a second entity having a presence on the online system;   predict a first measure of quality of the image by applying a quality prediction model to the image, wherein the quality prediction model is a machine-learning model trained to predict a measure of quality of an image based on the image and a set of attributes of the image;   determine that the first measure of quality exceeds a threshold measure of quality;   responsive to determining that the first measure of quality does not exceed the threshold measure of quality, modify the image to remove the second region of interest;   predict a second measure of quality of the modified image by applying the quality prediction model to the modified image;   determine that the second measure of quality exceeds the threshold measure of quality; and   responsive to determining that the second measure of quality exceeds the threshold measure of quality, add a set of data to the content item, wherein the set of data is associated with the first item and the first entity.   
     
     
         12 . The computer-readable medium of  claim 11 , wherein the set of data added to the content item comprises one or more selected from the group consisting of: a tag describing the item, a link to the page associated with the entity maintained in the online system, and a catalog of items associated with the entity maintained in the online system. 
     
     
         13 . The computer-readable medium of  claim 11 , further storing instructions that, when executed by a processor, cause the processor to:
 responsive to determining that the first measure of quality does not exceed the threshold, one or more selected from the group consisting of: crop a portion of the image, zoom into a portion of the image, increase a sharpness of the image, reduce an amount of noise characterizing the image, change one or more colors within the image, change a brightness of the image, and reduce an amount of distortion characterizing the image.   
     
     
         14 . The computer-readable medium of  claim 11 , wherein the quality prediction model is trained based on an additional set of images with different measures of quality. 
     
     
         15 . The computer-readable medium of  claim 14 , wherein the different measures of quality of the additional set of images are based at least in part on one or more selected from the group consisting of: a resolution of each of the additional set of images, a percentage of each of the additional set of images corresponding to an item associated with the entity, a number of items associated with one or more additional entities having a presence on the online system comprising each of the additional set of images, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises a content item comprising each of the additional set of images. 
     
     
         16 . The computer-readable medium of  claim 11 , wherein the predicted first measure of quality of the image is based at least in part on one or more selected from the group consisting of: a resolution of the image, a percentage of the image corresponding to the item associated with the entity, a number of items associated with the one or more additional entities having a presence on the online system comprising the image, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         17 . The computer-readable medium of  claim 16 , wherein the predicted first measure of quality is proportional to one or more of: the resolution of the image, the percentage of the image corresponding to the item associated with the entity, and the amount of user engagement with the link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         18 . The computer-readable medium of  claim 16 , wherein the predicted first measure of quality is inversely proportional to the number of items associated with the one or more additional entities having a presence on the online system comprising the image. 
     
     
         19 . The computer-readable medium of  claim 11 , wherein the predicted second measure of quality of the image is based at least in part on one or more selected from the group consisting of: a resolution of the image, a percentage of the image corresponding to the item associated with the entity, a number of items associated with the one or more additional entities having a presence on the online system comprising the image, and an amount of user engagement with a link to the page associated with the entity maintained in the online system, wherein the link comprises the content item received from the content-providing user. 
     
     
         20 . The computer-readable medium of  claim 11 , further storing instructions that cause the processor to:
 responsive to determining that the second predicted measure of quality exceeds the threshold measure of quality:
 identify one or more viewing users of the online system subscribing to content received from the content-providing user, wherein the one or more viewing users satisfy a set of targeting criteria received from the first entity; 
 determine a number of the one or more viewing users subscribing to the content received from the content-providing user; 
 determine a value of a performance metric associated with one or more content items received from the content-providing user, wherein the one or more content items are associated with one or more of:
 the one or more items associated with the first entity and a topic associated with the one or more items; and 
 
 compute an influencer score associated with the content-providing user based at least in part on one or more of: the number of the one or more viewing users subscribing to the content received from the content-providing user and the value of the performance metric.