Determining a similarity between tree related structures

1. A method for determining a similarity between tree related structures, said method comprising:

given a collection of trees, determining, at a computer system, according to an order a plurality of restricted top down model (RTDM) distances between a plurality of nodes residing on at least one tree related structure of said collection of trees, wherein said determining computes a distance between subtrees rooted at nodes for each pair of nodes and wherein said determining comprises traversing said at least one tree related structure exactly once and:

determining, at said computer system, a first RTDM distance measurement between nodes of a first node pair of said plurality of nodes, wherein said determining said first RTDM distance comprises comparing prior determined RTDM distance measurements with said first node pair, wherein said prior determined RTDM distance measurements relate to node pairs preceding said first node pair within said order;

comparing, at said computer system, said first RTDM distance measurement, said prior determined RTDM distance measurements and a second node pair of said plurality of nodes; and

based on said comparing, generating, at said computer system, a second RTDM distance measurement.

2. The method of claim 1 , wherein said determining according to an order at least one restricted top down model (RTDM) distance between a plurality of nodes residing on at least one tree related structure comprises:

determining according to a reversed pre-order said at least one restricted top down model (RTDM) distance between said plurality of nodes residing on said at least one tree related structure.

3. The method of claim 1 , further comprising:

storing, at said computer system, said first and second RTDM distance measurements in a distance matrix store; and

based on said first and second RTDM distance measurements stored in said distance matrix store, generating, at said computer system, a similarity measurement between nodes of at least one node pair of said plurality of node pairs to achieve a generated similarity measurement.

4. The method of claim 3 , further comprising:

clustering a first portion of said tree related structure based on said first and second RTDM distance measurements and said generated similarity measurement.

5. The method of claim 4 , wherein said generating a similarity measurement comprises:

determining a repetition within said tree related structure of a second portion of said tree related structure based on a size of said clustering.

6. The method of claim 5 , wherein said determining a repetition within said tree related structure of a second portion of said tree related structure based on a size of said clustering comprises:

determining a repetition of a structural arrangement on an area of a page; and

based on a predetermined amount of said repetition, determining a semantic importance of said structural arrangement.

7. The method of claim 4 , wherein said clustering a first portion of said tree related structure based on said first and second RTDM distance measurements comprises:

clustering said first portion of said tree related structure having a predetermined similarity measurement that represents a closeness between sub-portions of said first portion.

8. The method of claim 1 , wherein said determining a first RTDM distance between nodes of a first node pair of said plurality of nodes comprises:

determining a first RTDM distance between a first node from a first tree of a forest and a first node from a second tree of said forest.

9. The method of claim 8 , wherein said comparing said first RTDM distance measurement, said prior determined RTDM distance measurements and a second node pair of said plurality of nodes comprises:

comparing, at said computer system, said first RTDM distance measurement, said prior determined RTDM distance measurements and a second node from said first tree and a second node from said second tree.

10. The method of claim 1 , wherein said determining a first RTDM distance between nodes of a first node pair of said plurality of nodes comprises:

determining, at said computer system, a first RTDM distance between a first node from a first subtree of a tree and a first node from a second subtree of said tree.

11. The method of claim 10 , wherein said comparing said first RTDM distance measurement and said nodes of a second node pair of said plurality of nodes comprises:

comparing said first RTDM distance measurement, said prior determined RTDM distance measurements and a second node from said first subtree and a second node from said second subtree.

12. A non-transitory computer-readable storage medium comprising instructions stored thereon which, when executed by a computer system, cause said computer system to perform a method for determining a similarity between tree related structures, said method comprising:

given a collection of trees, determining according to an order a plurality of restricted top down model (RTDM) distance measurements between a plurality of nodes residing on at least one tree related structure of said collection of trees, wherein said determining computes a distance between subtrees rooted at nodes for each pair of nodes and wherein said determining comprises traversing said at least one tree related structure exactly once and:

determining a first RTDM distance measurement between nodes of a first node pair of said plurality of nodes, wherein said determining said first RTDM distance measurement comprises comparing prior determined RTDM distance measurements with said first node pair, wherein said prior determined RTDM distance measurements relate to node pairs preceding said first node pair within said order; wherein said first node pair comprises a first node from a first tree of a forest and a first node from a second tree of said forest;

comparing said first RTDM distance measurement, said prior determined RTDM distance measurements and a second node pair of said plurality of nodes; and

based on said comparing, generating a second RTDM distance measurement.

13. The non-transitory computer-readable storage medium of claim 12 , wherein said method further comprises:

determining according to a reversed order at least one RTDM distance between a plurality of nodes residing on at least one tree related structure.

14. The non-transitory computer-readable storage medium of claim 12 , wherein said method further comprises instructions for:

storing said first and second RTDM distance measurements in a distance matrix store; and

based on said first and second RTDM distance measurements stored in said distance matrix store, generating a similarity measurement between nodes of at least one node pair of said plurality of node pairs to achieve a generated similarity measurement.

15. The non-transitory computer-readable storage medium of claim 14 , wherein said method further comprises instructions for:

based on said first and second RTDM distance measurements and said generated similarity measurement, clustering a first portion of said forest.

16. The non-transitory computer-readable storage medium of claim 15 , wherein said method further comprises instructions for:

determining a repetition within said forest of a second portion of said forest based on a size of said clustering.

17. The non-transitory computer-readable storage medium of claim 16 , wherein said method further comprises instructions for:

determining a repetition of a structural arrangement on an area of a page; and

based on a predetermined amount of said repetition, determining a semantic importance of said structural arrangement.

18. The non-transitory computer-readable storage medium of claim 15 , wherein said method further comprises instructions for:

clustering said first portion of said forest close to each other based on similar first and second RTDM distance measurements.

19. A system comprising:

a first restricted top down model (RTDM) distance measurement determiner configured for determining according to an order a first RTDM distance measurement between nodes of a first node pair of a plurality of nodes of a tree related structure of a given collection of trees, wherein said determining computes a distance between subtrees rooted at nodes for each pair of nodes and wherein said determiner comprises traversing said tree related structure exactly once, wherein said first RTDM distance measurement is based on a comparison of prior determined RTDM distance measurements with said first node pair, wherein said prior determined RTDM distance measurements relate to node pairs preceding said first node pair within said order;

a comparer configured for comparing said first RTDM distance measurement, said prior determined RTDM distance measurements and said a second node pair of said plurality of nodes;

an RTDM distance measurement generator configured for generating a second RTDM distance measurement based on said comparing;

a distance matrix store configured for storing said first RTDM distance measurement, said prior determined RTDM distance measurements and said second RTDM distance measurement;

a similarity measurement generator configured for generating a similarity measurement between nodes of at least one node pair of said plurality of node pairs using said first and second RTDM distance measurements stored in said distance matrix store; and

a cluster generator configured for clustering a first portion of said tree related structure based on said first and second RTDM distance measurements and said similarity measurement.

20. The system of claim 19 , wherein said order is reversed.