Predicting popularity of electronic publications
AssigneeAmazon Technologies, Inc.
InventorsMichael Patrick Bacus, Sandesh S. Singh
Life of this patent
Estimating the term…
Abstract
A popularity prediction module receives an electronic media item and identifies a feature of the electronic media item. The popularity prediction module applies the feature of the electronic media item to a learned function, where the learned function is determined from a plurality of features from one or more other electronic media items that meet a popularity classification. The popularity prediction module predicts a popularity of the electronic media item based on the comparing, before providing the electronic media item to a user.
Claims18
A method comprising:
- receiving an electronic media item;
- identifying a feature of the electronic media item;
- applying, by a processor, the feature of the electronic media item to a learned function, wherein the learned function is determined using a plurality of features from one or more other electronic media items that are determined to have met a popularity threshold, wherein an output of the learned function to indicate a predicted relative popularity level of the electronic media item;
- classifying the electronic media item as at least one of popular or not popular in relation to the popularity threshold based on the predicted relative popularity level output from the learned function, before an initial release the electronic media item;
- and
- when the electronic media item is classified as popular in relation to the popularity threshold, enhancing the electronic media item by adding enriched content before the initial release of the electronic media item.
A system, comprising:
- a processor;
- and
- a memory coupled to the processor, the memory storing a popularity prediction module and an enriched content module, executable by the processor and from the memory, the popularity prediction module to:
- receive an electronic media item;
- predict a relative popularity level of the electronic media item using an output of a learned function created using a plurality of features from one or more other electronic media items that are determined to have met a popularity threshold;
- and
- compare the predicted relative popularity level of the electronic media item to the popularity threshold to classify the electronic media item as at least one of popular or not popular before an initial release of the electronic media item;
- and
- the enriched content module to:
- when the electronic media item is classified as popular in relation to the popularity threshold, enhance the electronic media item by adding enriched content before the initial release of the electronic media item.
A non-transitory computer-readable storage medium storing instructions which, when executed, cause a processor to perform operations comprising:
- predicting a relative popularity level of an electronic media item using an output of a learned function created using a plurality of features from one or more other electronic media items that are determined to have met a popularity threshold;
- classifying, by the processor, the electronic media item as at least one of popular or not popular in relation to the popularity threshold based on the predicted relative popularity level output from the learned function before an initial release of the electronic media item;
- and
- when the electronic media item is classified as popular in relation to the popularity threshold, enhancing the electronic media item by adding enriched content before the initial release of the electronic media item.
Dependencies are inferred from the claim text.