| 摘要 |
The distribution of content items (such as news items), in a news publishing platform is governed by a plurality of interrelated factors. These factors include publisher bias, trust bias, and user-specific bias (reflecting the user's reading history, the social clusters to which the user belongs, the user's location etc.). To model the relevance of each content item to a given user, a metric of the overall story strength is calculated in accordance with one or more of these biases. The content item is delivered to the target in accordance with the story strength metric. |
