Situational resource rating

1. A system comprising:

a storage device to store a respective semantic projection for each of a plurality of resources, for each of a plurality of sets of user interests and for each of a plurality of user contexts, each respective semantic projection conforming to a common semantic projection model; and

a processor to:

determine a similarity of a semantic projection of a set of interests of a first user to a semantic projection of a first resource includes:

determination of a similarity of a semantic projection of a set of static interests associated with the first user to the semantic projection of the first resource;

determination of a similarity of a semantic projection of a set of dynamic interests associated with the first user and a first context to the semantic projection of the resource, and

wherein the determination of the semantic interest of the first user for the first resource in the first context is based on the similarity of the semantic projection of the set of static interests associated with the first user to the semantic projection of the resource, and on the similarity of the semantic projection of the set of dynamic interests associated with the first user and the first context to the semantic projection of the resource;

determine a semantic interest of the first user for the first resource in the first context based on the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource; and

determine an interest measure associated with the first user, the first resource and the first context based on a rating prediction for the first user to the first resource in the first context and on the determined semantic interest of the first user for the first resource in the first context.

2. A system according to claim 1 , the processor further to:

determine one or more contexts similar to the first context based on a similarity between the semantic projection of the first context and respective one or more semantic projections of the one or more contexts; and

determine the rating prediction for the first user to the first resource in the first context based on a rating given by the first user to the first resource in the first context, a number of times the first resource has been suggested to the first user, and a number of times the first resource has been suggested to the first user in the one or more contexts similar to the first context.

3. A system according to claim 2 , the processor further to:

determine one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

4. A system according to claim 1 , the processor further to:

determine one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

5. A system according to claim 1 , the processor further to:

determine one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

6. A system according to claim 1 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

association of respective weights with each of first concepts of the semantic projection of the set of the interests of the first user based on the respective depths of each of the first concepts in the common semantic projection model;

association of respective weights with each of second concepts of the semantic projection of the first resource based on the respective depths of each of the second concepts in the common semantic projection model; and

determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource based on the respective weights of each of the first concepts and the respective weights of each of the second concepts.

7. A system according to claim 6 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

association of respective coefficients of explicitness with each of the first concepts of the semantic projection of the set of the interests of the first user based on whether each one of the first concepts is explicitly referenced;

association of respective coefficients of explicitness with each of the second concepts of the semantic projection of the first resource based on whether each one of the second concepts is explicitly referenced; and

determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource based on the respective weights and respective coefficients of explicitness of the first concepts and the respective weights and respective coefficients of explicitness of the second concepts.

8. A system according to claim 1 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

determination of a first branch of the semantic projection of the set of interests of the first user and a second branch of the semantic projection of the first resource which is comparable to the first branch;

determination of a first one or more sub-trees for the first branch, the first one or more sub-trees comprising first explicit concepts and parent concepts thereof;

determination of a second one or more sub-trees for the second branch, the second one or more sub-trees comprising second explicit concepts and parent concepts thereof; and

determination of a similarity between the first branch and the second branch based on a weighted average of a proximity of weights of concepts of the first one or more sub-trees and weights of corresponding concepts of the second one or more sub-trees.

9. A computer-implemented method comprising:

determining a similarity of a semantic projection of a set of interests of a first user to a semantic projection of a first resource, wherein the semantic projection of the set of interests of the first user and the semantic projection of the first resource conforms to a common semantic projection model, the determination including:

determining a similarity of a semantic projection of a set of static interests associated with the first user to the semantic projection of the first resource; and

determining a similarity of a semantic projection of a set of dynamic interests associated with the first user and a first context to the semantic projection of the resource;

wherein the determination of the semantic interest of the first user for the first resource in the first context is based on the similarity of the semantic projection of the set of static interests associated with the first user to the semantic projection of the resource, and on the similarity of the semantic projection of the set of dynamic interests associated with the first user and the first context to the semantic projection of the resource;

determining a semantic interest of the first user for the first resource in the first context based on the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource; and

determining an interest measure associated with the first user, the first resource and the first context based on a rating prediction for the first user to the first resource in the first context and on the determined semantic interest of the first user for the first resource in the first context.

10. A computer-implemented method according to claim 9 , further comprising:

determining one or more contexts similar to the first context based on a similarity between the semantic projection of the first context and respective one or more semantic projections of the one or more contexts; and

determining the rating prediction for the first user to the first resource in the first context based on a rating given by the first user to the first resource in the first context, a number of times the first resource has been suggested to the first user, and a number of times the first resource has been suggested to the first user in the one or more contexts similar to the first context.

11. A computer-implemented method according to claim 10 , further comprising:

determining one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

12. A computer-implemented method according to claim 9 , further comprising:

determining one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

13. A computer-implemented method according to claim 9 , further comprising:

determining one or more users similar to the first user based on a similarity between the semantic projection of the first user and respective one or more semantic projections of the one or more users,

wherein determination of the interest measure is based on the rating prediction for the first user to the first resource in the first context, on the determined semantic interest of the first user for the first resource in the first context, and on an average rating prediction for the one or more users to the first resource in the first context.

14. A computer-implemented method according to claim 9 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

associating respective weights with each of first concepts of the semantic projection of the set of the interests of the first user based on the respective depths of each of the first concepts in the common semantic projection model;

associating respective weights with each of second concepts of the semantic projection of the first resource based on the respective depths of each of the second concepts in the common semantic projection model; and

determining the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource based on the respective weights of each of the first concepts and the respective weights of each of the second concepts.

15. A computer-implemented method according to claim 14 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

associating respective coefficients of explicitness with each of the first concepts of the semantic projection of the set of the interests of the first user based on whether each one of the first concepts is explicitly referenced;

associating respective coefficients of explicitness with each of the second concepts of the semantic projection of the first resource based on whether each one of the second concepts is explicitly referenced; and

determining the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource based on the respective weights and respective coefficients of explicitness of the first concepts and the respective weights and respective coefficients of explicitness of the second concepts.

16. A computer-implemented method according to claim 9 , wherein determination of the similarity of the semantic projection of the set of the interests of the first user to the semantic projection of the first resource comprises:

determining a first branch of the semantic projection of the set of interests of the first user and a second branch of the semantic projection of the first resource which is comparable to the first branch;

determining a first one or more sub-trees for the first branch, the first one or more sub-trees comprising first explicit concepts and parent concepts thereof;

determining a second one or more sub-trees for the second branch, the second one or more sub-trees comprising second explicit concepts and parent concepts thereof; and

determining a similarity between the first branch and the second branch based on a weighted average of a proximity of weights of concepts of the first one or more sub-trees and weights of corresponding concepts of the second one or more sub-trees.