Training tree transducers
First Claim
Patent Images
1. A method, comprising:
- obtaining tree transducer information including input/output pair information and transducer information;
converting said input/output pair information and said transducer information into a set of values in a weighted tree grammar; and
using said weighted tree grammar to solve a problem that requires information from the input/output pair information and transducer information.
1 Assignment
0 Petitions
Accused Products
Abstract
Training using tree transducers is described. Given sample input/output pairs as training, and given a set of tree transducer rules, the information is combined to yield locally optimal weights for those rules. This combination is carried out by building a weighted derivation forest for each input/output pair and applying counting methods to those forests.
104 Citations
36 Claims
-
1. A method, comprising:
-
obtaining tree transducer information including input/output pair information and transducer information;
converting said input/output pair information and said transducer information into a set of values in a weighted tree grammar; and
using said weighted tree grammar to solve a problem that requires information from the input/output pair information and transducer information. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A training apparatus comprising:
-
an input part which includes tree transducer information that includes input samples and output samples, and includes transducer information;
a processor that processes said tree transducer information and said transducer information and produces an output table indicative of said tree transducer information and said transducer information, said output table including a set of values, and weights for said set of values. - View Dependent Claims (18, 19, 20, 21, 22, 23, 24, 25)
-
-
26. A computer readable storage medium containing a set of instructions to enable a computer to carry out a function, the set of instructions including instructions to:
-
obtain tree transducer information including at least an input tree, output information corresponding to said input tree, and transducer information;
convert said tree transducer information into a weighted set of values indicative of said tree transducer information and said transducer information; and
use said set of values to solve a problem. - View Dependent Claims (27, 28, 29)
-
-
30. A method comprising:
-
using a computer to obtain information in the form of a first-tree, second information corresponding to said first tree, and transducer information; and
using said computer to automatically distill information from said first tree, said from said second information, and said transducer information into a list of information in a specified tree grammar with weights associated with entries in the list and to produce locally optimal weights for said entries. - View Dependent Claims (31, 32, 33, 34, 35, 36)
-
Specification