Estimation of parameters for machine translation without in-domain parallel data

• US 9,652,453 B2
• Filed: 04/14/2014
• Issued: 05/16/2017
• Est. Priority Date: 04/14/2014
• Status: Expired due to Fees

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Mathur, Prashant, Venkatapathy, Sriram, Cancedda, Nicola
• Primary Examiner(s): JACKSON, JAKIEDA R
• Application Number:
  US14/252,032
• Publication Number:
  US 20150293908A1
• Time in Patent Office: 1,128 Days
• Field of Search: 704 2, 704 4
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/44 Statistical methods, e.g. probability models

Abstract

A system and method for estimating parameters for features of a translation scoring function for scoring candidate translations in a target domain are provided. Given a source language corpus for a target domain, a similarity measure is computed between the source corpus and a target domain multi-model, which may be a phrase table derived from phrase tables of comparative domains, weighted as a function of similarity with the source corpus. The parameters of the log-linear function for these comparative domains are known. A mapping function is learned between similarity measure and parameters of the scoring function for the comparative domains. Given the mapping function and the target corpus similarity measure, the parameters of the translation scoring function for the target domain are estimated. For parameters where a mapping function with a threshold correlation is not found, another method for obtaining the target domain parameter can be used.

Claims

1. A method for estimating parameters for features of a translation scoring function and for scoring candidate translations in a target domain comprising:

receiving a monolingual source corpus for a target domain and deriving n-gram counts from the monolingual source corpus or receiving n-gram counts derived only from the monolingual source corpus, the monolingual source corpus comprising sentences in a source language;

generating a multi-model for the target domain based on a phrase table for each of a set of comparative domains and a measure of similarity between the n-gram counts derived only from the source corpus for the target domain and the phrase tables for the comparative domains, each of the phrase tables storing a value for each of a set of features for each of a set of biphrases, the generated target domain multi-model being a weighted combination of two or more of the phrase tables for the comparative domains;

for the target domain, computing a measure of similarity between the monolingual source corpus and the target domain multi-model;

for each of a plurality of the comparative domains, computing a measure of similarity between a source corpus for the comparative domain and a respective comparative domain multi-model that is derived from phrase tables for others of the set of the comparative domains, each of the plurality of comparative domains being associated with parameters for at least some of the features of the translation scoring function;

estimating the parameters of the translation scoring function for the target domain based on the computed measure of similarity between the source corpus and the target domain multi-model, the computed measures of similarity for the comparative domains, and the parameters for the scoring function for the comparative domains; and

with a statistical machine translation component, scoring a translation with the translation scoring function,

wherein the generating of the target domain multi-model, computing the measure of similarity between the source corpus and the target domain multi-model, computing the measure of similarity between a source corpus for the comparative domains and the respective comparative domain multi-models, and the estimating the parameters for the translation scoring function are performed with a computer processor.

2. The method of claim 1, wherein the estimating of the parameters comprises:

learning a function which maps values of at least one parameter of the translation scoring function to the computed measures of similarity for the comparative domains; and

where the learned function indicates a correlation between the at least one parameter and the computed measures of similarity, estimating the at least one parameter for the target domain based on the learned function.

3. The method of claim 2, where the learned function is a linear regression function.

4. The method of claim 2, wherein when a predefined correlation is not found, estimating the at least one parameter for the translation scoring function based on the corresponding at least one parameter of one of the comparative domains that has a computed similarity with the respective comparative domain multi-model which is closest to the computed similarity with the target domain multi-model.

5. The method of claim 1, wherein each similarity measure is computed as a function of counts of n-grams of each of a plurality of sizes in the source corpus of the respective domain that are present in the phrase table or multi-model with which the similarity is being computed.

6. The method of claim 5, wherein each similarity measure may be computed as a function of

7. The method of claim 1, wherein the generating of the multi-model for the target domain comprises combining the phrase tables for the comparative domains in a weighted combination in which each of the comparative domain phrase tables is weighted as a function of the measure of similarity between the source corpus for the target domain and the comparative domain phrase table.

8. The method of claim 1, wherein the multi-model for a first of the comparative domains is generated by combining the phrase tables for others of the comparative domains in a weighted combination in which each of the other comparative domain phrase tables is weighted as a function of the measure of similarity between the source corpus for the first comparative domain and the other comparative domain phrase table.

9. The method of claim 1, wherein the method is performed without access to a parallel corpus in the target domain.

10. The method of claim 1, wherein the set of comparative domains comprises at least three comparative domains.

11. The method of claim 1, wherein the translation scoring function is a log-linear scoring function.

12. The method of claim 1, wherein the features of the translation scoring function include features selected from the group consisting of lexical features, phrasal features, reordering features, and language model features.

13. The method of claim 12, wherein the features of the translation scoring function include lexical features, phrasal features, reordering features, and at least one language model feature.

14. The method of claim 1, wherein each of the comparative domain phrase tables includes biphrase features for each of a set of biphrases, each biphrase including a source phrase and a corresponding target phrase, the biphrase features having been derived from a parallel corpus of source and target text strings.

15. A computer program product comprising a non-transitory recording medium storing instructions, which when executed on a computer, causes the computer to perform the method of claim 1.

16. A system for estimating parameters for features of a translation scoring function comprising memory which stores instructions for performing the method of claim 1 and a processor in communication with the memory for executing the instructions.

17. A machine translation system comprising memory which stores instructions for scoring translations of source text with a translation scoring function, the translation scoring function including parameters estimated by the method of claim 1, and a processor which executes the instructions.

18. The method of claim 1, wherein each of the comparative domain phrase tables includes at least 10,000 biphrases.

19. The method of claim 1, wherein the translation scoring function is a log-linear model of the general form:

score(t₁|s₁)=1/z exp(Σ

_m=1^Mλ

_mh_m(s₁,t₁)) 

 

(1),

where s₁ represents a source language text string, t₁ represents a candidate translation of the source string in the target language, h_m represents one of M features, λ

_m is a respective estimated parameter for the feature, and Z is an optional normalization constant.

20. The method of claim 19, wherein M is at least 9.

21. A system for estimating parameters for features of a translation scoring function for performing machine translation in a target domain comprising:

memory which stores a monolingual source corpus for a target domain or n-grams present the monolingual source corpus, the monolingual source corpus comprising sentences in a source language;

a similarity computation component which computes a measure of similarity between the target domain monolingual source corpus and a phrase table for each of a set of comparative domains by comparing n-grams present the monolingual source corpus and source language phrases in the phrase table;

a multi-model computation component which generates a multi-model for the target domain based on the phrase tables for the comparative domains and the computed measures of similarity, the generated target domain multi-model being a weighted combination of two or more of the phrase tables for the comparative domains;

the similarity computation component further computing, for the target domain, a measure of similarity between the source corpus and the target domain multi-model;

the similarity computation component further computing a measure of similarity for each of the comparative domains between a respective comparative domain source corpus and a respective comparative domain multi-model that is derived from phrase tables for others of the set of the comparative domains, each of the plurality of comparative domains being associated with parameters for at least some of the features of the translation scoring function;

a parameter computation component which estimates the parameters of the translation scoring function for the target domain based on the computed measure of similarity between the source corpus and the target domain multi-model, the computed measures of similarity for the comparative domains, and the parameters for the scoring function for the comparative domains;

a statistical machine translation component which scores translations of source text with the translation scoring function, at least some of the features of the translation scoring function being computed based on the target domain multi-model; and

a processor for implementing the similarity computation component, multi-model computation component, and parameter computation component.

22. A method for estimating parameters for features of a translation scoring function for scoring candidate translations in a target domain comprising:

for each of a plurality of parameters of the translation scoring function, learning a mapping function which maps a similarity measure to the parameter of the translation scoring function, the similarity measure being computed between a source corpus for one domain and a respective multi-model derived from phrase tables of other domains;

receiving a source corpus for a target domain;

generating a multi-model for the target domain based on phrase tables of comparative domains, the multi-model for the target domain being a phrase table which includes feature values for each of a set of such biphrases, the multi-model for the target domain being formed by combining at least two of the phrase tables of the comparative domains;

computing a measure of similarity between the target domain source corpus and the target domain multi-model;

based on the computed measure of similarity and the mapping functions, estimating the plurality of parameters for the translation scoring function for the target domain;

incorporating the translation scoring function into a statistical machine translation system,

wherein the learning of the mapping function, generating of the target domain multi-model, computing the measure of similarity, and the estimating of the set of parameters for the translation scoring function are performed with a computer processor.