Method and apparatus for recommending video

An apparatus for recommending a video, the apparatus comprising:

• at least one processor;
• and
• a memory storing instructions, the instructions when executed by the at least one processor, cause the at least one processor to perform operations, the operations comprising:
• generating, by using a pre-trained neural network, a content characteristic vector for each candidate video of all candidate videos based on a plurality of images in the each candidate video, wherein a component of the content characteristic vector generated for the each candidate video corresponds to a content characteristic comprising a text in an image in the each candidate video;
• finding a similar candidate video corresponding to each candidate video based on a similarity of content characteristic vectors, comprising:
• clustering content characteristic vectors of all the candidate videos to obtain a plurality of clustering results;
• for the content characteristic vector of each candidate video in each clustering result of the plurality of clustering results, determining the similarity between the content characteristic vector of the candidate video and the content characteristic vector of each of other candidate videos in the clustering result, sorting the other candidate videos in a descending order based on similarities of the content characteristic vectors of the other candidate videos to the content characteristic vector of the candidate video, defining a top preset number of the sorted candidate videos as cluster similar videos of the candidate video in the clustering result, and defining the cluster similar video corresponding to each candidate video in the clustering result as the similar candidate video corresponding to each candidate video, each component of the content characteristic vector of the candidate video corresponding to a content characteristic of the candidate video, the content characteristic of the candidate video comprising text in an image in the candidate video;
• finding a recommended video corresponding to a target video from all the candidate videos based on similarities of content characteristics of the videos, the target video being a video to be played on a terminal of a user;
• and
• sending play information of the recommended video corresponding to the target video to the terminal of the user.

A non-transitory computer storage medium storing a computer program, the computer program when executed by one or more processors, causes the one or more processors to perform operations, the operations comprising:

• generating, by using a pre-trained neural network, a content characteristic vector for each candidate video of all candidate videos based on a plurality of images in the each candidate video, wherein a component of the content characteristic vector generated for the each candidate video corresponds to a content characteristic comprising a text in an image in the each candidate video;
• finding a similar candidate video corresponding to each candidate video based on a similarity of content characteristic vectors, comprising:
• clustering content characteristic vectors of all the candidate videos to obtain a plurality of clustering results;
• for the content characteristic vector of each candidate video in each clustering result of the plurality of clustering results, determining the similarity between the content characteristic vector of the candidate video and the content characteristic vector of each of other candidate videos in the clustering result, sorting the other candidate videos in a descending order based on similarities of the content characteristic vectors of the other candidate videos to the content characteristic vector of the candidate video, defining a top preset number of the sorted candidate videos as cluster similar videos of the candidate video in the clustering result, and defining the cluster similar video corresponding to each candidate video in the clustering result as the similar candidate video corresponding to each candidate video, each component of the content characteristic vector of the candidate video corresponding to a content characteristic of the candidate video, the content characteristic of the candidate video comprising text in an image in the candidate video;
• finding a recommended video corresponding to a target video from all the candidate videos based on similarities of content characteristics of the videos, the target video being a video to be played on a terminal of a user;
• and
• sending play information of the recommended video corresponding to the target video to the terminal of the user.