Recommendation system based on a trendsetter inference

摘要

A trend setting parameter, which is a degree of trend setting shown by a user, is generated for each of several users. This degree of trending displayed by the user is an indication of how well the user identifies trends for objects (e.g., consumes objects) before the objects become popular. Object consumption by users with high trend setting parameters is then used to identify objects that are expected to become popular after a time delay. For a given user, a further user is identified who has a high trend setting value (and who is also referred to as a trendsetter) and who has a high affinity for the given user (for example a similar behavior or similar properties when consuming objects). Recommendations are provided to the given user based on objects that have been consumed by the trendsetter before the objects become popular.