Weighting attributes for media content-based recommendations

摘要

Techniques for automatically assigning weights to attributes of media content are disclosed, in part, based on how many users actually viewed or heard the content, as well as how many users liked the content or otherwise specified a preference to have. The content objects may be any type of audio or visual media content, such as songs, videos, or movies, as well as written content, such as books, articles, journals, advertisements, or magazines. A first similarity score is determined based on a similarity between user preferences for content objects. A second similarity score is determined based on a similarity between one or more shared attributes of the content objects. Attributes are assigned to these attributes, which represent the number of users who have consumed the corresponding content. Next, each of the attributes is assigned weights based on the first and second similarity characteristics, for example, using regression techniques with linear equations.