Methods and apparatuses for content preparation and/or selection

1. A machine implemented method for content preparation or selection, the method performed by a computing system configured to operate a content preparation or selection module, the method comprising:

receiving, as part of the content preparation or selection, by the content preparation or selection module, a representation of a concept; and

for a first content item (N) that is included in a content hierarchy that includes a plurality of content items:

determining, for the content preparation or selection, by the content preparation or selection module, a score of topical relevance of the first content item (N) to the concept;

wherein the determining includes a computation of propagation of influence among the plurality of content items of the content hierarchy where influence is a capacity of an nth content item to affect scores of topical relevance of other content items to the concept; and

wherein the computation of propagation of influence among the plurality of content items of the content hierarchy is subject to constraints that bound the number of operations performed by the computing system to perform the computation with a chosen constant.

2. The method of claim 1 , wherein the constraints include a constraint that limits a number of siblings of a given node in the content hierarchy considered in the computation of the propagation of influence among content items of the content hierarchy.

3. The method of claim 2 , wherein the constraints include a constraint that limits a maximum number of terminal content item hits recognized.

4. The method of claim 2 , wherein the constraints include a constraint that requires a minimum term-based score for a terminal content item hit to be recognized.

5. The method of claim 1 , wherein the concept is a search expression, and the constraints include a constraint that limits a number of nodes of the content hierarchy visited to compute the propagation of influence for a content item to m*d*k nodes,

where m is a number of terminal content nodes that count as active matches for the search expression,

d is a maximum number of ancestors of a node, added to 1, and

k is a determined maximum number of neighbors of a node, including the node as its own neighbor.

6. The method of claim 5 , wherein determining the score of topical relevance of the first content item (N) to the concept includes calculating, initially, term-based scores of terminal content items of the content hierarchy.

7. The method of claim 6 , wherein calculating, initially, term-based scores of terminal content items of the content hierarchy comprises calculating, initially, term-based scores of terminal content items that qualify as active matches.

8. The method of claim 1 , wherein the concept is a search expression, and wherein the method further comprises pruning the search expression according to a function derived from frequencies of documents of the content items of the content hierarchy.

9. The method of claim 8 , wherein the constraints include a constraint that requires a minimum inverse-document-frequency for a term to remain in a pruned search expression.

10. The method of claim 1 , wherein the concept is a search expression, and wherein the receiving further comprises receiving the constant that bounds the number of operations from a user issuing the search expression.

11. The method of claim 1 , wherein the computation of propagation of influence among content items of the content hierarchy is performed to realize transitivity of relevance among the content items of the content hierarchy.

12. An apparatus for content preparation or selection, comprising:

one or more processors; and

a content preparation or selection module operated by the one or more processors to:

receive, as part of the content preparation or selection, a representation of a concept; and

for a first content item (N) that is included in a content hierarchy that includes a plurality of content items:

determine, for the content preparation or selection, a score of topical relevance of the first content item (N) to the concept;

wherein the determining includes a computation of propagation of influence among the plurality A of content items of the content hierarchy where influence is a capacity of an nth content item to affect scores of topical relevance of other content items to the concept; and

wherein the computation of propagation of influence among the plurality of content items of the content hierarchy is subject to constraints that bound the number of operations performed by the computing system to perform the computation with a chosen constant.

13. The method of claim 1 , wherein the content preparation or selection module is operatively coupled to a content storage or access system comprising a computer-readable memory, and further comprising accessing from the content storage or access system, by the content preparation or selection module, the content hierarchy that includes a plurality of content items.

14. The method of claim 1 , wherein the constraints that bound the number of operations performed by the computing system allow the computation of propagation of influence among content items of the content hierarchy to be completed in a time frame consistent with a real time response to a user search request or recommendation generation.