Trust propagation through both explicit and implicit social networks
The present invention is directed towards systems and methods for trust propagation. The method according to one embodiment comprises calculating a first feature vector for a first user, calculating a second feature for a second user and comparing the first feature vector with the second feature vector to calculate a similarity value. A determination is made as to whether the similarity value falls within a threshold. If the similarity value falls within the threshold, a relationship is recorded between the first user and the second user in a first user profile and a second user profile.
1. A method for propagating trust to create an implicit social network, the method comprising:
storing a content item in an inverted index;
receiving profile information from at least one user, the profile information relating to actions by the user to the content item including an annotation submitted by the user for the content item, the annotation including descriptive text generated by the user;
recording the profile information to a stream search queue when an update threshold is not exceeded, whether the update threshold is exceeded determined based on a last update of the inverted index;
storing the contents of the stream queue to the inverted index when the update threshold is exceeded;
calculating a first feature vector for a first user based on first user profile information;
calculating a second feature vector for a second user based on second user profile information;
calculating a similarity value using the first feature vector and the second feature vector indicating if the actions by the first user to the content item are similar to the actions by the second user to the content item;
when the similarity value falls within a feature threshold, recording a relationship between the first user and the second user in a first user profile and a second user profile, the feature threshold based in part on features used to calculate the first feature vector and the second feature vector;
in response to a search request by the first user, determining a search result set, the search result set includes the content item, and profile information in the inverted index and the stream search queue; ranking the search result set based on the second user profile information related to actions by the second user to the content item when the similarity value falls within the features threshold;
wherein ranking comprises:
selecting the content item from the result set, selecting the second user profile, calculating a usage total for the content item when an indicia of the content item appears in the second user profile and assigning a trust rank value to the content item based on the usage total; and
wherein the trust rank value is further determined based on:
TrustRank=Σ i r i *TrustRank/Σ i r i ,
where TrustRank(f) is the trust rank value for content item f, TrustRank(u i ) is the trust rank of user u i and r i is the rating of the content item f by user u i .
2. The method of claim 1 wherein calculating the first feature vector further comprises calculating on the basis of a tag by the first user for a content item.
3. The method of claim 1 wherein calculating the first feature vector further comprises calculating on the basis of a save by the first user of a content item.
4. The method of claim 1 wherein determining comprises determining if the first user and second user are related in accordance with an implicit social network.
5. The method of claim 1 wherein the first feature vector indicates that the first user saved a content item and the second feature vector indicates that the second user saved the content item.
6. The method of claim 1 wherein the first feature vector indicates that the first user tagged a content item and the second feature vector indicates that the second user tagged the content item.
7. The method of claim 1 comprising:
calculating a third feature vector for a third user;
comparing the first feature vector with the third feature vector to calculate a similarity value;
recording a relationship between the first user and the third user in the first user profile and a third user profile based upon the similarity value, wherein the similarity value falls within the feature threshold;
retrieving the third user profile based on the recorded relationship between the first user and the third user; and
ranking the search result set based on third user profile information relating to actions by the third user to the content item.
8. The method of claim 7 wherein determining comprises determining if the first user and the third user are related in accordance with an implicit social network.