A method and system for determining similar users may include receiving information for a source user, the information including at least one topic and a user value for each topic, where the value represents how strongly the user is associated with that topic. Similarity scores may be generated based on a value for each topic for the source user and a value for the same topic for each user in a set of users, where each user in the set of users is associated with a value for each topic. One or more similar users may be selected based on the generated similarity scores, and one or more of the selected users may be output.