Method and system for translating sentence between languages based on semantic structure of the sentence

• US 9,817,818 B2
• Filed: 05/21/2012
• Issued: 11/14/2017
• Est. Priority Date: 10/10/2006
• Status: Expired due to Fees

• Current Assignee: ABBYY Production LLC (ABBYY Software)
• Original Assignee: ABBYY Production LLC (ABBYY Software)
• Inventors: Anisimovich, Konstantin, Selegey, Vladimir, Zuev, Konstantin
• Primary Examiner(s): ROBERTS, SHAUN A
• Application Number:
  US13/477,021
• Publication Number:
  US 20120232883A1
• Time in Patent Office: 2,003 Days
• Field of Search: 704 2- 4, 704 7- 9
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 40/56 Natural language generation

Abstract

A method and computer system for translating sentences between languages from an intermediate language-independent semantic representation is provided. On the basis of comprehensive understanding about languages and semantics, exhaustive linguistic descriptions are used to analyze sentences, to build syntactic structures and language independent semantic structures and representations, and to synthesize one or more sentences in a natural or artificial language. A computer system is also provided to analyze and synthesize various linguistic structures and to perform translation of a wide spectrum of various sentence types. As result, a generalized data structure, such as a semantic structure, is generated from a sentence of an input language and can be transformed into a natural sentence expressing its meaning correctly in an output language. The method and computer system can be applied to in automated abstracting, machine translation, natural language processing, control systems, Internet information retrieval, etc.

Claims

1. A method of translating a source sentence in a source language into an output sentence in an output language, the method comprising:

performing, by a computer system, a lexico-morphological analysis of the source sentence to build a lexico-morphological structure representing the source sentence;

performing, using the lexico-morphological structure, a rough syntactic analysis of the source sentence to generate a graph of generalized constituents representing multiple possible syntactic relationships between words of the source sentence;

performing, using the graph of generalized constituents, a precise syntactic analysis of the source sentence to generate a plurality of syntactic trees representing the source sentence, wherein each syntactic tree comprises a subgraph of the graph of generalized constituents;

determine, for the plurality of syntactic trees, values of a syntactic tree rating reflecting a degree of correspondence of a filler of a surface slot of a constituent of the syntactic structure to a restriction associated with the surface slot;

selecting, based on the values of the syntactic tree rating, one or more syntactic trees among the plurality of the syntactic trees;

generating, based on the selected syntactic trees, a syntactic structure representing the source sentence;

using the syntactic structure to generate a language-independent semantic structure representing the source sentence; and

using the language-independent semantic structure to generate the output sentence.

2. The method of claim 1, wherein the lexico-morphological structure comprises one or more lexical meaning-grammatical value pairs corresponding to each lexeme of the source sentence.

3. The method of claim 1, wherein performing the syntactic analysis comprises generating a surface model for each lexical meaning-grammatical value pair of a lexico-morphological structure of the source sentence.

4. The method of claim 1, wherein the restriction comprises at least one of:

a grammatical restriction or a semantic restriction associated with the surface slot.

5. The method of claim 1, wherein the syntactic tree rating reflects a degree of compatibility of lexical meanings of two or more lexemes of the source sentence.

6. The method of claim 1, wherein generating the output sentence further comprises producing a syntactic structure of the output sentence.

7. The method of claim 1, further comprising:

responsive to failing to assign non-tree links to the selected syntactic tree, selecting a second syntactic tree among the plurality of the syntactic trees.

8. A non-transitory computer-readable storage medium encoded with instructions for causing a computing system to:

perform a lexico-morphological analysis of a source sentence to build a lexico-morphological structure representing the source sentence;

perform, using the lexico-morphological structure, a syntactic analysis of the source sentence to generate a plurality of syntactic trees representing the source sentence;

determine, for the plurality of syntactic trees, values of a syntactic tree rating reflecting a degree of correspondence of a filler of a surface slot of a constituent of the syntactic structure to a restriction associated with the surface slot;

select, based on the values of the syntactic tree rating, a first syntactic tree among the plurality of the syntactic trees;

responsive to failing to assign non-tree links to the first syntactic tree, selecting a second syntactic tree among the plurality of the syntactic trees;

generate, based on the second syntactic tree, a syntactic structure representing the source sentence;

using the syntactic structure to generate a language-independent semantic structure representing the source sentence; and

using the language-independent semantic structure to generate an output sentence in a specified output language.

9. The non-transitory computer-readable storage medium of claim 8, wherein the lexico-morphological structure comprises one or more lexical meaning-grammatical value pairs corresponding to each lexeme of the source sentence.

10. The non-transitory computer-readable storage medium of claim 8, wherein performing the syntactic analysis comprises generating a surface model for each lexical meaning-grammatical value pair of a lexico-morphological structure of the source sentence.

11. The non-transitory computer-readable storage medium of claim 8, wherein the syntactic tree rating reflects a degree of correspondence of a filler of a surface slot of a constituent of the syntactic structure to at least one of:

a grammatical restriction or a semantic restriction associated with the surface slot.

12. The non-transitory computer-readable storage medium of claim 8, wherein the syntactic tree rating reflects a degree of compatibility of lexical meanings of two or more lexemes of the source sentence.

13. The non-transitory computer-readable storage medium of claim 8, wherein generating the output sentence further comprises producing a syntactic structure of the output sentence.

14. The non-transitory computer-readable storage medium of claim 8, wherein performing the syntactic analysis of the source sentence further comprises:

performing, using the lexico-morphological structure, a rough syntactic analysis of the source sentence to generate a graph of generalized constituents representing multiple possible syntactic relationships between words of the source sentence; and

performing, using the graph of generalized constituents, a precise syntactic analysis of the source sentence to generate a plurality of syntactic trees representing the source sentence, wherein each syntactic tree comprises a subgraph of the graph of generalized constituents.

15. A computer system for translating a source sentence from a source language into an output sentence in an output language, the computer system comprising:

a memory;

a processor coupled to the memory, the processor configured to;

perform a lexico-morphological analysis of the source sentence to build a lexico-morphological structure representing the source sentence;

perform, using the lexico-morphological structure, a syntactic analysis of the source sentence to generate a plurality of syntactic trees representing the source sentence;

determine, for the plurality of syntactic trees, values of a syntactic tree rating reflecting a degree of correspondence of a filler of a surface slot of a first constituent of the syntactic structure to a restriction associated with the surface slot;

select, based on the values of the syntactic tree rating, a first syntactic tree among the plurality of the syntactic trees;

responsive to failing to assign non-tree links to the first syntactic tree, select a second syntactic tree among the plurality of the syntactic trees;

generate, based on the second syntactic tree, a syntactic structure representing the source sentence;

use the syntactic structure of the source sentence to generate a language-independent semantic structure representing the source sentence; and

use the language-independent semantic structure to generate the output sentence.

16. The computer system of claim 15, wherein the lexico-morphological structure comprises one or more lexical meaning-grammatical value pairs corresponding to each lexeme of the source sentence.

17. The computer system of claim 15, wherein performing the syntactic analysis comprises generating a surface model for each lexical meaning-grammatical value pair of a lexico-morphological structure of the source sentence.

18. The computer system of claim 15, wherein the syntactic tree rating reflects a degree of correspondence of a filler of a surface slot of a constituent of the syntactic structure to at least one of:

a grammatical restriction or a semantic restriction associated with the surface slot.

19. The computer system of claim 15, wherein the syntactic tree rating reflects a degree of correspondence of a lexical meaning associated with a lexeme of the source sentence to a pragmatic description associated with the source sentence.

20. The computer system of claim 15, wherein the syntactic tree rating reflects a degree of compatibility of lexical meanings of two or more lexemes of the source sentence.

21. The computer system of claim 15, wherein generating the output sentence further comprises producing a syntactic structure of the output sentence.

22. The computer system of claim 15, wherein performing the syntactic analysis of the source sentence further comprises:

performing, using the lexico-morphological structure, a rough syntactic analysis of the source sentence to generate a graph of generalized constituents representing multiple possible syntactic relationships between words of the source sentence; and

performing, using the graph of generalized constituents, a precise syntactic analysis of the source sentence to generate a plurality of syntactic trees representing the source sentence, wherein each syntactic tree comprises a subgraph of the graph of generalized constituents.