| 主权项 |
1. A video-revenue prediction tool comprising:
a memory that is configured to store a set of monetized videos and historical revenue for each video in the set of monetized videos, wherein the memory is further configured to store a set of non-monetized videos, each monetized video having revenue history; a feature extractor that is configured to extract one or more features from each monetized video in the set of monetized videos and to combine the features extracted from each monetized video into a training-set feature vector, wherein the feature extractor is further configured to extract one or more features from each non-monetized video in the set of non-monetized videos and to combine the features extracted from each non-monetized video into a non-monetized video feature vector for each non-monetized video, each non-monetized video having no revenue history; a model-estimating module that is configured to map the training-set feature vector to the historical revenue for each monetized video for the training-set feature vector; a video-revenue predictor that is configured to yield a predicted revenue for each non-monetized video in the set of non-monetized videos based at least in part upon the training-set feature vector and the mapping between the training-set feature vector and the historical revenue; and an up-loader selector that is configured to determine whether to invite an uploader of a non-monetized video to monetize the up-loaded non-monetized video based at least in part on the predicted revenue of the up-loaded non-monetized video, and wherein, in response to a determination to invite the up-loader of the non-monetized video to monetize the up-loaded non-monetized video, the up-loader selector is further configured to invite the up-loader to monetize the up-loaded non-monetized video. |
