Hierarchical topic relevance
View Patent ↗Determining the relevance of a page to a topic in a hierarchy is disclosed. A plurality of paths that include arrivals at the page is determined. A proportion of the paths that include relevant arrivals at the page is determined. And, the relevance of the page is determined based at least in part on the proportion.
1. A computer-implemented method of determining the relevance of a page to a topic in a hierarchy comprising:
receiving a seed set of nodes, wherein the seed set of nodes includes a first seed node and a second seed node, and wherein the first seed node and the second seed node are both associated with the topic in the hierarchy;
determining a plurality of paths of a plurality of surfers that include arrivals at the page;
determining the relevance for each of the plurality of surfers as either being relevant or irrelevant based at least in part on the surfers' behaviors corresponding to the plurality of paths; and
determining the relevance of the page to the topic in the hierarchy, at least in part by determining a ratio of (1) a measure of paths included in the plurality that include visiting at least one of either the first seed node or the second seed node prior to visiting the page and (2) a measure of the plurality of paths and (3) the relevance of the surfers corresponding to the plurality of paths.
2. The method of claim 1 wherein the topic hierarchy is a tree.
3. The method of claim 1 wherein the topic hierarchy is a directed acyclic graph.
4. The method of claim 1 wherein the topic hierarchy is arranged such that children of a node represent subtopics of a topic with which the node is associated.
5. The method of claim 1 wherein there is a seed set of pages associated with every topic in the hierarchy.
6. The method of claim 1 wherein a seed set for a topic includes the seed sets of all its subtopics.
7. The method of claim 1 wherein a pair of topics is unrelated if neither is an ancestor of the other.
8. The method of claim 1 wherein associated with each topic is also an excluded set of pages.
9. The method of claim 8 wherein the excluded set contains pages that are known not to belong to the topic.
10. The method of claim 1 wherein at least some values for an excluded set of pages are pinned and not updated.
11. The method of claim 1 further comprising determining a value that expresses a page's contribution to topic drift.
12. A system for determining the relevance of a page to a topic in a hierarchy comprising:
a processor, configured to:
receive a seed set of nodes, wherein the seed set of nodes includes a first seed node and a second seed node, and wherein the first seed node and the second seed node are both associated with the topic in the hierarchy;
determine a plurality of paths of a plurality of surfers that include arrivals at the page; determine the relevance for each of the plurality of surfers as either being relevant or irrelevant based at least in part on the surfers' behaviors corresponding to the plurality of paths;
determine the relevance of the page to the topic in the, at least in part by determining a ratio of (1) a measure of paths included in the plurality that include visiting at least one of either the first seed node or the second seed node prior to visiting the page and (2) a measure of the plurality of paths and (3) the relevance of the surfers corresponding to the plurality of paths; and
a memory, coupled with the processor, wherein the memory is configured to provide the processor with instructions.
13. The system of claim 12 wherein the topic hierarchy is a tree.
14. The system of claim 12 wherein the topic hierarchy is a directed acyclic graph.
15. The system of claim 12 wherein the topic hierarchy is arranged such that children of a node represent subtopics of a topic with which the node is associated.
16. The system of claim 12 wherein there is a seed set of pages associated with every topic in the hierarchy.
17. The system of claim 12 wherein a seed set for a topic includes the seed sets of all its subtopics.
18. The system of claim 12 wherein a pair of topics is unrelated if neither is an ancestor of the other.
19. The system of claim 12 wherein associated with each topic is also an excluded set of pages.
20. A computer program product for determining the relevance of a page to a topic in a hierarchy, the computer program product being embodied in a computer readable storage medium and comprising computer instructions for:
receiving a seed set of nodes, wherein the seed set of nodes includes a first seed node and a second seed node, and wherein the first seed node and the second seed node are both associated with the topic in the hierarchy;
determining a plurality of paths of a plurality of surfers that include arrivals at the page;
determining the relevance for each of the plurality of surfers as either being relevant or irrelevant based at least in part on the surfers' behaviors corresponding to the plurality of paths; and
determining the relevance of the page to the topic in the hierarchy, at least in part by determining a ratio of (1) a measure of paths included in the plurality that include visiting at least one of either the first seed node or the second seed node prior to visiting the page and (2) a measure of the plurality of paths and (3) the relevance of the surfers corresponding to the plurality of paths.